In [8]:
rc_data, rc_st_data, rc_st_tp_data = creating_time_series(data_path, ts_path)
data is loaded!
import pandas as pd
import os
import numpy as np
import utils
import time
from statsmodels.tsa.stattools import acf, pacf
from statsmodels.graphics.tsaplots import plot_acf, plot_pacf
from pmdarima import auto_arima # for determining ARIMA orders
import matplotlib.ticker as ticker
formatter = ticker.StrMethodFormatter('{x:,.0f}')
from statsmodels.tsa.arima_model import ARMA,ARMAResults,ARIMA,ARIMAResults
from statsmodels.tsa.statespace.sarimax import SARIMAX
from statsmodels.graphics.tsaplots import plot_acf,plot_pacf # for determining (p,q) orders
from pmdarima import auto_arima # for determining ARIMA orders
from sklearn.metrics import r2_score, mean_squared_error
from sklearn.preprocessing import MinMaxScaler
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import LSTM
from keras.preprocessing.sequence import TimeseriesGenerator
import prophet
from prophet import Prophet
from prophet.diagnostics import performance_metrics
from prophet.diagnostics import cross_validation
from prophet.plot import plot_cross_validation_metric
# dir(utils)
from utils import creating_time_series
from utils import load_single_timeseries
from utils import timeseries_eda
from utils import check_adfuller
from utils import model_evaluation
input_path = f"{os.getcwd()}\\input_data\\"
output_path = "C:\\Users\\DKici\\Desktop\\Metrolinx\\"
ts_path = f"{os.getcwd()}\\ts_data\\"
data_file = "interview_dataset.xlsx"
figures_path = f'{output_path}\\figures\\'
data_path = input_path + data_file
'C:\\Users\\DKici\\Desktop\\Metrolinx\\'
Single Time Series for each rail corridor and station
rc_data, rc_st_data, rc_st_tp_data = creating_time_series(data_path, ts_path)
data is loaded!
holidays = pd.read_csv(f'{input_path}/holidays.csv')
holidays.columns= holidays.columns.str.lower()
holidays = holidays.iloc[:,:-2]
# print(holidays)
rc_st_data
['Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data', 'Metrolinx_Barrie_AURORA GO_Data']
my_evaluation = pd.DataFrame({'model': [],
'ts': [],
'mse':[],
'rmse':[],
'R2':[],
'time (min)' : []
})
for st in rc_st_data:
print(f"================={st}==================")
ts = f"{st}.csv"
ts_data_path = ts_path + ts
print(ts_data_path)
data, data_all, data_regressors = load_single_timeseries(ts_data_path)
print(data)
timeseries_eda(data)
print(f"--------------")
check_adfuller(data)
fig, axes = plt.subplots(1,2,figsize=(16,3), dpi= 100)
plot_acf(data, lags=50, ax=axes[0]);
plot_pacf(data, lags=50, ax=axes[1]);
data_diff = data - data.shift()
plt.figure(figsize=(22,10))
plt.plot(data_diff)
plt.title("Differencing method")
plt.xlabel("Date")
plt.ylabel("Differencing Mean Ridership")
plt.show()
fig, axes = plt.subplots(1,2,figsize=(16,3), dpi= 100)
plot_acf(data_diff, lags=50, ax=axes[0]);
plot_pacf(data_diff, lags=50, ax=axes[1]);
train = data.iloc[:-365,:] # 3 years (75%)
test = data.iloc[-365:,:] # 1 year (25%)
len(train),len(test)
##################################### MODEL 1 - ARIMA #####################################
# Start timer
start_time = time.perf_counter()
arima_model = auto_arima(train, start_p=0, start_q=0,
max_p=2, max_q=2, m=6,
d=None, trace=True,
error_action='ignore', # we don't want to know if an order does not work
suppress_warnings=True, # we don't want convergence warnings
stepwise=True) # set to stepwise
arima_model.summary()
# End timer
end_time = time.perf_counter()
# Calculate elapsed time
elapsed_time = end_time - start_time
print("Elapsed time: ", elapsed_time)
# Forecast
prediction = arima_model.predict(n_periods=len(test))
# prediction
prediction_series = pd.Series(prediction,index=test.index)
fig, ax = plt.subplots(1, 1, figsize=(15, 5))
ax.plot(data.ridership)
ax.plot(prediction_series)
mse, rmse, r2 = model_evaluation(test['ridership'], prediction_series, "ARIMA", st)
mse_score.append(mse)
rmse_score.append(rmse)
r2_score_.append(r2)
r = my_evaluation.shape[0]
my_evaluation.loc[r] = ['ARIMA', st, mse, rmse, r2, elapsed_time/60]
##################################### LSTM ####################################
scaler = MinMaxScaler()
scaler.fit(train.values.reshape(-1,1))
scaled_train = scaler.transform(train.values.reshape(-1,1))
scaled_test = scaler.transform(test.values.reshape(-1,1))
# Let's redefine to get 3 months back and then predict the next month out
n_input = 30
n_features = 1
generator = TimeseriesGenerator(scaled_train, scaled_train, length=n_input, batch_size=1)
# define model
model = Sequential()
model.add(LSTM(100, activation='relu', input_shape=(n_input, n_features)))
model.add(Dense(1))
model.compile(optimizer='adam', loss='mse')
model.summary()
# Start timer
start_time = time.perf_counter()
# fit model
model.fit_generator(generator,epochs=100)
# End timer
end_time = time.perf_counter()
# Calculate elapsed time
elapsed_time = end_time - start_time
print("Elapsed time: ", elapsed_time)
loss_per_epoch = model.history.history['loss']
plt.plot(range(len(loss_per_epoch)),loss_per_epoch)
first_eval_batch = scaled_train[-30:]
first_eval_batch = first_eval_batch.reshape((1, n_input, n_features))
model.predict(first_eval_batch)
test_predictions = []
first_eval_batch = scaled_train[-n_input:]
current_batch = first_eval_batch.reshape((1, n_input, n_features))
for i in range(len(test)):
# get prediction 1 time stamp ahead ([0] is for grabbing just the number instead of [array])
current_pred = model.predict(current_batch)[0]
# store prediction
test_predictions.append(current_pred)
# update batch to now include prediction and drop first value
current_batch = np.append(current_batch[:,1:,:],[[current_pred]],axis=1)
# Inverse Transformations and Compare
true_predictions = scaler.inverse_transform(test_predictions)
test = pd.DataFrame(test)
test['Predictions'] = true_predictions
test.plot(figsize=(12,8))
mse, rmse, r2 = model_evaluation(test['ridership'], test['Predictions'], "LSTM", st)
mse_score.append(mse)
rmse_score.append(rmse)
r2_score_.append(r2)
r = my_evaluation.shape[0]
my_evaluation.loc[r] = ['LSTM', st, mse, rmse, r2, elapsed_time/60]
# # save the model
# model.save('ridership_lstm_model.h5')
# # load the model back
# from keras.models import load_model
# new_model = load_model('ridership_lstm_model.h5')
# new_model.summary()
############################## PROPHET ##############################
df = data.merge(data_regressors, on ='date')
df = df.reset_index()
df = df.rename(columns = {'ridership':'y', 'date':'ds'})
print(df.head())
##### Model Development with Cross-Validation
# Start timer
start_time = time.perf_counter()
print(f" \n ########## Cross validation for {st} ######### \n ")
m = Prophet(holidays=holidays
,holidays_prior_scale=0.25
,changepoint_prior_scale=0.01
,seasonality_mode='additive'
,weekly_seasonality=True)
m.add_regressor('trips' , prior_scale=0.5, mode='additive')
m.fit(df)
m_cv = cross_validation(m, initial='730 days', period='180 days', horizon = '365 days')
# End timer
end_time = time.perf_counter()
# Calculate elapsed time
elapsed_time = end_time - start_time
print("Elapsed time: ", elapsed_time)
metrics = performance_metrics(m_cv)
# print(f" \n Ridership - CV-metrics \n")
# print("metrics are:",metrics)
fig = plot_cross_validation_metric(m_cv, metric='rmse')
fig = plot_cross_validation_metric(df_cv=m_cv, metric='mae', rolling_window=0.1)
print(f" \n Ridership {st}- CV-metrics mean \n")
print(metrics.mean())
data_regressors = data_regressors.reset_index()
data_futures = data_regressors.copy()
data_futures = data_futures.rename(columns={'date':'ds'})
print(data_futures)
forecast = m.predict(data_futures)
forecast.to_csv(f"C:\\Users\\DKici\\Desktop\\Metrolinx\\forecasts\\Prophet_{st}_Ridership_forecast.csv")
f = m.plot_components(forecast).savefig(f'{figures_path}/Prophet_{st}_Components_Ridership.jpeg');
from prophet.plot import add_changepoints_to_plot
fig = m.plot(forecast, xlabel='Date', ylabel='Ridership')
a = add_changepoints_to_plot(fig.gca(), m, forecast)
ax = fig.gca()
ax.set_title(f"Prophet {st} Ridership", size=24)
fig.savefig(f'{figures_path}/Prophet_{st}_change_points_Ridership.jpeg');
mse, rmse, r2 = model_evaluation(test['ridership'], forecast.iloc[-365:,-1], "Prophet", st)
mse_score.append(mse)
rmse_score.append(rmse)
r2_score_.append(r2)
r = my_evaluation.shape[0]
my_evaluation.loc[r] = ['Prophet', st, mse, rmse, r2, elapsed_time/60]
# print(f" \n Ridership- regressor coefficients {regressor_coefficients}")
# print(regressor_coefficients(m))
# ########### Save the model ###########
# #In Python, models should not be saved with pickle;
# #the Stan backend attached to the model object will not pickle well,
# #and will produce issues under certain versions of Python.
# #Instead, you should use the built-in serialization functions to serialize the model to json:
# # from prophet.serialize import model_to_json, model_from_json
# # with open(f'{path}/models/serialized_model_Ridership.json', 'w') as fout:
# # fout.write(model_to_json(m)) # Save model
=================Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-02 174 3
2019-01-03 4218 7
2019-01-04 4434 7
2019-01-05 3886 7
2019-01-06 346 3
Data:
ridership
date
2019-01-02 174
2019-01-03 4218
2019-01-04 4434
2019-01-05 3886
2019-01-06 346
Data Regressors:
trips
date
2019-01-02 3
2019-01-03 7
2019-01-04 7
2019-01-05 7
2019-01-06 3
ridership
date
2019-01-02 174
2019-01-03 4218
2019-01-04 4434
2019-01-05 3886
2019-01-06 346
... ...
2022-12-28 840
2022-12-29 962
2022-12-30 930
2022-12-31 752
2023-01-01 641
[1461 rows x 1 columns]
--------------
Test statistic: -1.9171147753895854
p-value: 0.3240531073979981
Critical Values: {'1%': -3.4349056408696814, '5%': -2.863552005375758, '10%': -2.5678411776130114}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=19311.599, Time=0.41 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=19440.081, Time=0.01 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=19417.431, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=19002.063, Time=0.63 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=19438.082, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=19421.908, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=19003.680, Time=1.02 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=19003.096, Time=2.03 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=19416.427, Time=0.16 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=18891.335, Time=2.17 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=18752.603, Time=2.55 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=18716.550, Time=1.63 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=18886.685, Time=1.43 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=19149.453, Time=0.89 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=18713.404, Time=1.86 sec ARIMA(1,1,1)(1,0,1)[6] intercept : AIC=18924.982, Time=1.15 sec ARIMA(1,1,1)(2,0,0)[6] intercept : AIC=18848.708, Time=1.71 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=18818.520, Time=3.32 sec ARIMA(1,1,1)(1,0,0)[6] intercept : AIC=19051.669, Time=0.77 sec ARIMA(1,1,1)(1,0,2)[6] intercept : AIC=18805.094, Time=2.55 sec ARIMA(1,1,0)(2,0,1)[6] intercept : AIC=19109.613, Time=1.26 sec ARIMA(2,1,1)(2,0,1)[6] intercept : AIC=18623.453, Time=2.55 sec ARIMA(2,1,1)(1,0,1)[6] intercept : AIC=18777.477, Time=1.21 sec ARIMA(2,1,1)(2,0,0)[6] intercept : AIC=18812.947, Time=2.37 sec ARIMA(2,1,1)(2,0,2)[6] intercept : AIC=18618.265, Time=3.37 sec ARIMA(2,1,1)(1,0,2)[6] intercept : AIC=18690.091, Time=2.74 sec ARIMA(2,1,0)(2,0,2)[6] intercept : AIC=inf, Time=3.01 sec ARIMA(2,1,2)(2,0,2)[6] intercept : AIC=18306.948, Time=3.28 sec ARIMA(2,1,2)(1,0,2)[6] intercept : AIC=18474.839, Time=3.44 sec ARIMA(2,1,2)(2,0,1)[6] intercept : AIC=18666.448, Time=3.27 sec ARIMA(2,1,2)(1,0,1)[6] intercept : AIC=19010.342, Time=1.41 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=18803.285, Time=3.11 sec ARIMA(2,1,2)(2,0,2)[6] : AIC=18301.815, Time=2.73 sec ARIMA(2,1,2)(1,0,2)[6] : AIC=18472.024, Time=2.79 sec ARIMA(2,1,2)(2,0,1)[6] : AIC=18572.375, Time=2.00 sec ARIMA(2,1,2)(1,0,1)[6] : AIC=18605.744, Time=1.05 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=18734.176, Time=2.59 sec ARIMA(2,1,1)(2,0,2)[6] : AIC=18615.953, Time=2.73 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=18771.605, Time=2.52 sec Best model: ARIMA(2,1,2)(2,0,2)[6] Total fit time: 71.947 seconds Elapsed time: 72.0073553000002
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data ARIMA MSE Error: 2055632.928
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data ARIMA RMSE Error: 1433.747861
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data ARIMA R square: -0.7895198693
Model: "sequential_3"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_3 (LSTM) (None, 100) 40800
dense_3 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1066/1066 [==============================] - 5s 4ms/step - loss: 0.0551 Epoch 2/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0404 Epoch 3/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0221 Epoch 4/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0183 Epoch 5/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0143 Epoch 6/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0109 Epoch 7/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0098 Epoch 8/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0086 Epoch 9/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0085 Epoch 10/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0084 Epoch 11/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0085 Epoch 12/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0078 Epoch 13/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0078 Epoch 14/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0075 Epoch 15/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0074 Epoch 16/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0069 Epoch 17/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0073 Epoch 18/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0071 Epoch 19/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0070 Epoch 20/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0069 Epoch 21/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0067 Epoch 22/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0066 Epoch 23/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 24/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0065 Epoch 25/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 26/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0066 Epoch 27/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 28/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0066 Epoch 29/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 30/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 31/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.1111 Epoch 32/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0063 Epoch 33/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0057 Epoch 34/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 35/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 36/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0060 Epoch 37/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 38/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0060 Epoch 39/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0102 Epoch 40/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0060 Epoch 41/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 42/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 43/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 44/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 45/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 46/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 47/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 48/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 49/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0057 Epoch 50/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0068 Epoch 51/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0056 Epoch 52/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 53/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0057 Epoch 54/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0056 Epoch 55/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0057 Epoch 56/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0056 Epoch 57/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0055 Epoch 58/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 59/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 60/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 61/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 62/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 63/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 64/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 65/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 66/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 67/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 68/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 69/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 70/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 71/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 72/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 73/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 74/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 75/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 76/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 77/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 78/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 79/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 80/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 81/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 82/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 83/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 84/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 85/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 86/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 87/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 88/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 89/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 90/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 91/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 92/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 93/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 94/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 95/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 96/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 97/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 98/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 99/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 100/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Elapsed time: 416.45354510000016
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data LSTM MSE Error: 3450533.431
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data LSTM RMSE Error: 1857.561151
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data LSTM R square: -2.003842782
ds y trips
0 2019-01-02 174 3
1 2019-01-03 4218 7
2 2019-01-04 4434 7
3 2019-01-05 3886 7
4 2019-01-06 346 3
########## Cross validation for Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data #########
16:43:23 - cmdstanpy - INFO - Chain [1] start processing 16:43:23 - cmdstanpy - INFO - Chain [1] done processing
16:43:23 - cmdstanpy - INFO - Chain [1] start processing 16:43:24 - cmdstanpy - INFO - Chain [1] done processing 16:43:25 - cmdstanpy - INFO - Chain [1] start processing 16:43:25 - cmdstanpy - INFO - Chain [1] done processing 16:43:26 - cmdstanpy - INFO - Chain [1] start processing 16:43:26 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.038891500000318
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 5658895.3934
rmse 2348.268976
mae 1912.505077
mdape 1.672824
smape 1.470741
coverage 0.453241
dtype: object
ds trips
0 2019-01-02 3
1 2019-01-03 7
2 2019-01-04 7
3 2019-01-05 7
4 2019-01-06 3
... ... ...
1456 2022-12-28 6
1457 2022-12-29 9
1458 2022-12-30 9
1459 2022-12-31 9
1460 2023-01-01 6
[1461 rows x 2 columns]
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data Prophet MSE Error: 550028.0818
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data Prophet RMSE Error: 741.6387812
Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data Prophet R square: 0.5211760974
=================Metrolinx_Barrie_AURORA GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Barrie_AURORA GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-02 379 10
2019-01-03 373 8
2019-01-04 569 8
2019-01-05 413 8
2019-01-06 533 10
Data:
ridership
date
2019-01-02 379
2019-01-03 373
2019-01-04 569
2019-01-05 413
2019-01-06 533
Data Regressors:
trips
date
2019-01-02 10
2019-01-03 8
2019-01-04 8
2019-01-05 8
2019-01-06 10
ridership
date
2019-01-02 379
2019-01-03 373
2019-01-04 569
2019-01-05 413
2019-01-06 533
... ...
2022-12-28 585
2022-12-29 482
2022-12-30 500
2022-12-31 492
2023-01-01 492
[1461 rows x 1 columns]
--------------
Test statistic: -1.4073622643313042
p-value: 0.578741511289505
Critical Values: {'1%': -3.4349024693573584, '5%': -2.8635506057382325, '10%': -2.5678404322793846}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=13290.757, Time=0.24 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=13292.464, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=13202.135, Time=0.10 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=12936.100, Time=0.27 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=13290.471, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=12938.560, Time=0.11 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=12938.008, Time=0.42 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=12937.492, Time=0.81 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=12936.402, Time=0.26 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=12913.669, Time=1.51 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=inf, Time=3.06 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=12909.506, Time=1.10 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=12936.289, Time=0.68 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=13292.756, Time=0.55 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=12887.572, Time=2.63 sec ARIMA(1,1,1)(1,0,1)[6] intercept : AIC=inf, Time=1.31 sec ARIMA(1,1,1)(2,0,0)[6] intercept : AIC=12889.961, Time=2.08 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=inf, Time=3.98 sec ARIMA(1,1,1)(1,0,0)[6] intercept : AIC=12888.161, Time=0.74 sec ARIMA(1,1,1)(1,0,2)[6] intercept : AIC=inf, Time=2.86 sec ARIMA(1,1,0)(2,0,1)[6] intercept : AIC=13174.423, Time=1.22 sec ARIMA(2,1,1)(2,0,1)[6] intercept : AIC=12869.126, Time=2.30 sec ARIMA(2,1,1)(1,0,1)[6] intercept : AIC=12867.140, Time=1.14 sec ARIMA(2,1,1)(0,0,1)[6] intercept : AIC=12865.161, Time=0.70 sec ARIMA(2,1,1)(0,0,0)[6] intercept : AIC=12863.336, Time=0.33 sec ARIMA(2,1,1)(1,0,0)[6] intercept : AIC=12865.161, Time=0.64 sec ARIMA(1,1,1)(0,0,0)[6] intercept : AIC=12887.438, Time=0.47 sec ARIMA(2,1,0)(0,0,0)[6] intercept : AIC=13104.565, Time=0.05 sec ARIMA(2,1,2)(0,0,0)[6] intercept : AIC=12845.649, Time=0.57 sec ARIMA(2,1,2)(1,0,0)[6] intercept : AIC=12844.859, Time=1.00 sec ARIMA(2,1,2)(2,0,0)[6] intercept : AIC=12845.776, Time=2.17 sec ARIMA(2,1,2)(1,0,1)[6] intercept : AIC=12847.189, Time=1.59 sec ARIMA(2,1,2)(0,0,1)[6] intercept : AIC=12844.520, Time=1.20 sec ARIMA(2,1,2)(0,0,2)[6] intercept : AIC=12844.203, Time=2.84 sec ARIMA(2,1,2)(1,0,2)[6] intercept : AIC=12848.510, Time=3.26 sec ARIMA(1,1,2)(0,0,2)[6] intercept : AIC=12878.537, Time=0.93 sec ARIMA(2,1,1)(0,0,2)[6] intercept : AIC=12867.161, Time=1.66 sec ARIMA(1,1,1)(0,0,2)[6] intercept : AIC=12890.123, Time=2.46 sec ARIMA(2,1,2)(0,0,2)[6] : AIC=12842.551, Time=0.86 sec ARIMA(2,1,2)(0,0,1)[6] : AIC=12842.842, Time=0.48 sec ARIMA(2,1,2)(1,0,2)[6] : AIC=inf, Time=2.16 sec ARIMA(2,1,2)(1,0,1)[6] : AIC=12826.745, Time=0.75 sec ARIMA(2,1,2)(1,0,0)[6] : AIC=12843.183, Time=0.51 sec ARIMA(2,1,2)(2,0,1)[6] : AIC=12847.167, Time=1.25 sec ARIMA(2,1,2)(0,0,0)[6] : AIC=12843.997, Time=0.29 sec ARIMA(2,1,2)(2,0,0)[6] : AIC=12844.100, Time=1.13 sec ARIMA(2,1,2)(2,0,2)[6] : AIC=inf, Time=3.03 sec ARIMA(1,1,2)(1,0,1)[6] : AIC=12877.066, Time=0.31 sec ARIMA(2,1,1)(1,0,1)[6] : AIC=12855.072, Time=0.54 sec ARIMA(1,1,1)(1,0,1)[6] : AIC=12880.517, Time=0.92 sec Best model: ARIMA(2,1,2)(1,0,1)[6] Total fit time: 59.546 seconds Elapsed time: 59.616995500000485
Metrolinx_Barrie_AURORA GO_Data ARIMA MSE Error: 84279.99205
Metrolinx_Barrie_AURORA GO_Data ARIMA RMSE Error: 290.3101652
Metrolinx_Barrie_AURORA GO_Data ARIMA R square: -0.8707348856
Model: "sequential_4"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_4 (LSTM) (None, 100) 40800
dense_4 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1066/1066 [==============================] - 5s 4ms/step - loss: 0.0057 Epoch 2/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 3/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 4/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 5/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 6/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 7/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 8/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 9/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 10/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 11/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 12/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 13/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 14/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 15/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 16/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 17/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 18/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 19/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 20/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 21/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 22/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 23/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 24/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 25/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 26/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 27/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 28/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 29/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 30/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 31/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 32/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 33/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 34/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 35/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 36/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 37/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 38/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 39/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 40/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 41/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 42/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 43/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 44/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 45/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 46/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 47/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 48/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 49/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 50/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 51/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 52/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 53/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 54/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 55/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 56/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 57/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 58/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 59/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 60/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 61/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 62/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 63/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 64/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 65/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 66/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 67/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0029 Epoch 68/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 69/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 70/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 71/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 72/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 73/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 74/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 75/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 76/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 77/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 78/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 79/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 80/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 81/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 82/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 83/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 84/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 85/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 86/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 87/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 88/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 89/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 90/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 91/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 92/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 93/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 94/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 95/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0024 Epoch 96/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 97/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0024 Epoch 98/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0024 Epoch 99/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0024 Epoch 100/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0024 Elapsed time: 418.95170270000017
Metrolinx_Barrie_AURORA GO_Data LSTM MSE Error: 159561.7203
Metrolinx_Barrie_AURORA GO_Data LSTM RMSE Error: 399.4517746
Metrolinx_Barrie_AURORA GO_Data LSTM R square: -2.541738309
ds y trips
0 2019-01-02 379 10
1 2019-01-03 373 8
2 2019-01-04 569 8
3 2019-01-05 413 8
4 2019-01-06 533 10
########## Cross validation for Metrolinx_Barrie_AURORA GO_Data #########
16:51:41 - cmdstanpy - INFO - Chain [1] start processing 16:51:41 - cmdstanpy - INFO - Chain [1] done processing
16:51:42 - cmdstanpy - INFO - Chain [1] start processing 16:51:42 - cmdstanpy - INFO - Chain [1] done processing 16:51:43 - cmdstanpy - INFO - Chain [1] start processing 16:51:44 - cmdstanpy - INFO - Chain [1] done processing 16:51:45 - cmdstanpy - INFO - Chain [1] start processing 16:51:45 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.732848099999501
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Barrie_AURORA GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 71910.381578
rmse 255.276064
mae 217.328029
mdape 0.750609
smape 1.275425
coverage 0.315068
dtype: object
ds trips
0 2019-01-02 10
1 2019-01-03 8
2 2019-01-04 8
3 2019-01-05 8
4 2019-01-06 10
... ... ...
1456 2022-12-28 9
1457 2022-12-29 6
1458 2022-12-30 5
1459 2022-12-31 6
1460 2023-01-01 9
[1461 rows x 2 columns]
Metrolinx_Barrie_AURORA GO_Data Prophet MSE Error: 11656.93362
Metrolinx_Barrie_AURORA GO_Data Prophet RMSE Error: 107.9672803
Metrolinx_Barrie_AURORA GO_Data Prophet R square: 0.7412549308
=================Metrolinx_Barrie_UNION STATION_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Barrie_UNION STATION_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-02 596 12
2019-01-03 5102 18
2019-01-04 5636 18
2019-01-05 4837 18
2019-01-06 670 12
Data:
ridership
date
2019-01-02 596
2019-01-03 5102
2019-01-04 5636
2019-01-05 4837
2019-01-06 670
Data Regressors:
trips
date
2019-01-02 12
2019-01-03 18
2019-01-04 18
2019-01-05 18
2019-01-06 12
ridership
date
2019-01-02 596
2019-01-03 5102
2019-01-04 5636
2019-01-05 4837
2019-01-06 670
... ...
2022-12-28 933
2022-12-29 1269
2022-12-30 1355
2022-12-31 1210
2023-01-01 655
[1461 rows x 1 columns]
--------------
Test statistic: -1.8924925082508435
p-value: 0.3356010722555509
Critical Values: {'1%': -3.4349056408696814, '5%': -2.863552005375758, '10%': -2.5678411776130114}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=19673.385, Time=0.30 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=19820.357, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=19800.397, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=19361.675, Time=0.62 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=19818.357, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=19803.321, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=19363.650, Time=0.91 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=19523.673, Time=0.98 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=19799.739, Time=0.16 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=19250.255, Time=2.19 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=19209.221, Time=2.55 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=19100.780, Time=1.75 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=19259.291, Time=1.41 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=19523.828, Time=1.07 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=19095.643, Time=2.06 sec ARIMA(1,1,1)(1,0,1)[6] intercept : AIC=19287.730, Time=1.26 sec ARIMA(1,1,1)(2,0,0)[6] intercept : AIC=19221.770, Time=1.84 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=19203.055, Time=3.92 sec ARIMA(1,1,1)(1,0,0)[6] intercept : AIC=19433.436, Time=0.82 sec ARIMA(1,1,1)(1,0,2)[6] intercept : AIC=19164.956, Time=2.59 sec ARIMA(1,1,0)(2,0,1)[6] intercept : AIC=19486.498, Time=1.36 sec ARIMA(2,1,1)(2,0,1)[6] intercept : AIC=19081.711, Time=2.70 sec ARIMA(2,1,1)(1,0,1)[6] intercept : AIC=19153.381, Time=1.29 sec ARIMA(2,1,1)(2,0,0)[6] intercept : AIC=19110.240, Time=2.55 sec ARIMA(2,1,1)(2,0,2)[6] intercept : AIC=18991.468, Time=3.23 sec ARIMA(2,1,1)(1,0,2)[6] intercept : AIC=19060.888, Time=2.87 sec ARIMA(2,1,0)(2,0,2)[6] intercept : AIC=18945.677, Time=3.08 sec ARIMA(2,1,0)(1,0,2)[6] intercept : AIC=19398.902, Time=1.33 sec ARIMA(2,1,0)(2,0,1)[6] intercept : AIC=19116.711, Time=2.00 sec ARIMA(2,1,0)(1,0,1)[6] intercept : AIC=19532.592, Time=0.50 sec ARIMA(1,1,0)(2,0,2)[6] intercept : AIC=19457.235, Time=1.48 sec ARIMA(2,1,0)(2,0,2)[6] : AIC=19029.265, Time=2.39 sec Best model: ARIMA(2,1,0)(2,0,2)[6] intercept Total fit time: 49.426 seconds Elapsed time: 49.47169579999991
Metrolinx_Barrie_UNION STATION_Data ARIMA MSE Error: 3088077.236
Metrolinx_Barrie_UNION STATION_Data ARIMA RMSE Error: 1757.292587
Metrolinx_Barrie_UNION STATION_Data ARIMA R square: -1.436000175
Model: "sequential_5"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_5 (LSTM) (None, 100) 40800
dense_5 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1066/1066 [==============================] - 5s 4ms/step - loss: 0.0524 Epoch 2/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0464 Epoch 3/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0427 Epoch 4/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0281 Epoch 5/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0189 Epoch 6/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0159 Epoch 7/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0118 Epoch 8/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0101 Epoch 9/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0093 Epoch 10/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0087 Epoch 11/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0088 Epoch 12/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0080 Epoch 13/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0076 Epoch 14/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0079 Epoch 15/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0075 Epoch 16/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0077 Epoch 17/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0071 Epoch 18/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0070 Epoch 19/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0077 Epoch 20/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0065 Epoch 21/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0066 Epoch 22/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 23/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0066 Epoch 24/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 25/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 26/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 27/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0063 Epoch 28/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0067 Epoch 29/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0057 Epoch 30/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 31/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0060 Epoch 32/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 33/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0057 Epoch 34/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 35/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 36/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0060 Epoch 37/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 38/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0056 Epoch 39/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0054 Epoch 40/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0057 Epoch 41/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0054 Epoch 42/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0056 Epoch 43/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0055 Epoch 44/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 45/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 46/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0068 Epoch 47/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 48/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 49/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 50/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0055 Epoch 51/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 52/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0056 Epoch 53/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 54/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 55/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 56/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 57/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 58/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 59/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0050 Epoch 60/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 61/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0046 Epoch 62/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0051 Epoch 63/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 64/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 65/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 66/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 67/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 68/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 69/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 70/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0049 Epoch 71/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 72/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 73/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 74/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 75/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 76/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 77/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 78/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0041 Epoch 79/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 80/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 81/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 82/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 83/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 84/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 85/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 86/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 87/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 88/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 89/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0043 Epoch 90/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 91/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0039 Epoch 92/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 93/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 94/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 95/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 96/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 97/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 98/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 99/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0073 Epoch 100/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Elapsed time: 429.678492
Metrolinx_Barrie_UNION STATION_Data LSTM MSE Error: 5683112.553
Metrolinx_Barrie_UNION STATION_Data LSTM RMSE Error: 2383.927967
Metrolinx_Barrie_UNION STATION_Data LSTM R square: -3.483068951
ds y trips
0 2019-01-02 596 12
1 2019-01-03 5102 18
2 2019-01-04 5636 18
3 2019-01-05 4837 18
4 2019-01-06 670 12
########## Cross validation for Metrolinx_Barrie_UNION STATION_Data #########
17:00:00 - cmdstanpy - INFO - Chain [1] start processing 17:00:01 - cmdstanpy - INFO - Chain [1] done processing
17:00:01 - cmdstanpy - INFO - Chain [1] start processing 17:00:01 - cmdstanpy - INFO - Chain [1] done processing 17:00:02 - cmdstanpy - INFO - Chain [1] start processing 17:00:03 - cmdstanpy - INFO - Chain [1] done processing 17:00:04 - cmdstanpy - INFO - Chain [1] start processing 17:00:04 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.5169683000003715
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Barrie_UNION STATION_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 7223319.37399
rmse 2672.099812
mae 2084.966655
mdape 1.275367
smape 1.475115
coverage 0.502886
dtype: object
ds trips
0 2019-01-02 12
1 2019-01-03 18
2 2019-01-04 18
3 2019-01-05 18
4 2019-01-06 12
... ... ...
1456 2022-12-28 15
1457 2022-12-29 16
1458 2022-12-30 16
1459 2022-12-31 16
1460 2023-01-01 15
[1461 rows x 2 columns]
Metrolinx_Barrie_UNION STATION_Data Prophet MSE Error: 809423.2696
Metrolinx_Barrie_UNION STATION_Data Prophet RMSE Error: 899.6795372
Metrolinx_Barrie_UNION STATION_Data Prophet R square: 0.3614941998
=================Metrolinx_Lakeshore East_UNION STATION_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Lakeshore East_UNION STATION_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-02 26 1
2019-01-03 3397 14
2019-01-04 3615 14
2019-01-05 3130 14
2019-01-06 226 1
Data:
ridership
date
2019-01-02 26
2019-01-03 3397
2019-01-04 3615
2019-01-05 3130
2019-01-06 226
Data Regressors:
trips
date
2019-01-02 1
2019-01-03 14
2019-01-04 14
2019-01-05 14
2019-01-06 1
ridership
date
2019-01-02 26
2019-01-03 3397
2019-01-04 3615
2019-01-05 3130
2019-01-06 226
... ...
2022-12-28 3132
2022-12-29 4238
2022-12-30 4604
2022-12-31 4421
2023-01-01 1860
[1461 rows x 1 columns]
--------------
Test statistic: -1.9277800750981222
p-value: 0.31910743113489337
Critical Values: {'1%': -3.4349056408696814, '5%': -2.863552005375758, '10%': -2.5678411776130114}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=21072.210, Time=0.17 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=21191.334, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=21174.707, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=20836.335, Time=0.56 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=21189.334, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=21177.526, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=20838.273, Time=0.93 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=20962.040, Time=0.40 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=21174.544, Time=0.15 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=20733.679, Time=2.14 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=20652.435, Time=1.65 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=20598.169, Time=1.77 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=20737.998, Time=1.56 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=20959.176, Time=0.58 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=20575.919, Time=1.98 sec ARIMA(1,1,1)(1,0,1)[6] intercept : AIC=20729.300, Time=1.12 sec ARIMA(1,1,1)(2,0,0)[6] intercept : AIC=20667.908, Time=1.77 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=20636.376, Time=2.97 sec ARIMA(1,1,1)(1,0,0)[6] intercept : AIC=20848.833, Time=0.80 sec ARIMA(1,1,1)(1,0,2)[6] intercept : AIC=20623.962, Time=2.54 sec ARIMA(1,1,0)(2,0,1)[6] intercept : AIC=20899.558, Time=1.43 sec ARIMA(2,1,1)(2,0,1)[6] intercept : AIC=20517.912, Time=2.88 sec ARIMA(2,1,1)(1,0,1)[6] intercept : AIC=20653.540, Time=0.51 sec ARIMA(2,1,1)(2,0,0)[6] intercept : AIC=20566.363, Time=2.47 sec ARIMA(2,1,1)(2,0,2)[6] intercept : AIC=20460.044, Time=3.15 sec ARIMA(2,1,1)(1,0,2)[6] intercept : AIC=20539.863, Time=2.89 sec ARIMA(2,1,0)(2,0,2)[6] intercept : AIC=20478.147, Time=1.87 sec ARIMA(2,1,2)(2,0,2)[6] intercept : AIC=20232.253, Time=3.40 sec ARIMA(2,1,2)(1,0,2)[6] intercept : AIC=20696.565, Time=2.92 sec ARIMA(2,1,2)(2,0,1)[6] intercept : AIC=20547.527, Time=3.36 sec ARIMA(2,1,2)(1,0,1)[6] intercept : AIC=20621.029, Time=1.39 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=20562.157, Time=3.23 sec ARIMA(2,1,2)(2,0,2)[6] : AIC=20230.195, Time=2.90 sec ARIMA(2,1,2)(1,0,2)[6] : AIC=20554.495, Time=2.43 sec ARIMA(2,1,2)(2,0,1)[6] : AIC=20464.030, Time=2.75 sec ARIMA(2,1,2)(1,0,1)[6] : AIC=20613.656, Time=1.21 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=20560.124, Time=2.61 sec ARIMA(2,1,1)(2,0,2)[6] : AIC=20458.042, Time=2.75 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=20635.399, Time=2.67 sec Best model: ARIMA(2,1,2)(2,0,2)[6] Total fit time: 68.127 seconds Elapsed time: 68.1879614999998
Metrolinx_Lakeshore East_UNION STATION_Data ARIMA MSE Error: 20869683.08
Metrolinx_Lakeshore East_UNION STATION_Data ARIMA RMSE Error: 4568.334825
Metrolinx_Lakeshore East_UNION STATION_Data ARIMA R square: -2.063682612
Model: "sequential_6"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_6 (LSTM) (None, 100) 40800
dense_6 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1066/1066 [==============================] - 5s 4ms/step - loss: 0.0315 Epoch 2/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0275 Epoch 3/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0253 Epoch 4/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0202 Epoch 5/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0141 Epoch 6/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0123 Epoch 7/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0114 Epoch 8/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0099 Epoch 9/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0082 Epoch 10/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0065 Epoch 11/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0065 Epoch 12/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0060 Epoch 13/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 14/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0055 Epoch 15/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0054 Epoch 16/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 17/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0050 Epoch 18/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 19/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 20/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 21/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 22/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 23/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 24/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 25/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 26/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 27/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 28/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 29/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 30/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 31/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 32/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 33/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 34/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 35/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 36/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 37/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 38/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 39/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 40/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 41/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 42/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 43/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 44/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 45/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 46/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 47/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 48/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 49/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 50/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 51/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 52/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 53/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 54/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 55/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 56/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 57/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 58/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 59/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 60/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 61/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 62/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 63/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 64/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 65/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 66/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 67/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 68/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 69/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 70/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 71/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 72/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 73/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 74/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 75/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 76/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 77/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0031 Epoch 78/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 79/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 80/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 81/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 82/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 83/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 84/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 85/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 86/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 87/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 88/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 89/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 90/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 91/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 92/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 93/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 94/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 95/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 96/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 97/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 98/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 99/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 100/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Elapsed time: 422.8175124999998
Metrolinx_Lakeshore East_UNION STATION_Data LSTM MSE Error: 35106920.73
Metrolinx_Lakeshore East_UNION STATION_Data LSTM RMSE Error: 5925.109344
Metrolinx_Lakeshore East_UNION STATION_Data LSTM R square: -4.153718061
ds y trips
0 2019-01-02 26 1
1 2019-01-03 3397 14
2 2019-01-04 3615 14
3 2019-01-05 3130 14
4 2019-01-06 226 1
########## Cross validation for Metrolinx_Lakeshore East_UNION STATION_Data #########
17:08:31 - cmdstanpy - INFO - Chain [1] start processing 17:08:32 - cmdstanpy - INFO - Chain [1] done processing
17:08:33 - cmdstanpy - INFO - Chain [1] start processing 17:08:33 - cmdstanpy - INFO - Chain [1] done processing 17:08:34 - cmdstanpy - INFO - Chain [1] start processing 17:08:34 - cmdstanpy - INFO - Chain [1] done processing 17:08:35 - cmdstanpy - INFO - Chain [1] start processing 17:08:36 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.740547200000037
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Lakeshore East_UNION STATION_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 37625790.99809
rmse 6048.105489
mae 5156.497156
mdape 1.220829
smape 1.334697
coverage 0.263229
dtype: object
ds trips
0 2019-01-02 1
1 2019-01-03 14
2 2019-01-04 14
3 2019-01-05 14
4 2019-01-06 1
... ... ...
1456 2022-12-28 37
1457 2022-12-29 41
1458 2022-12-30 41
1459 2022-12-31 41
1460 2023-01-01 35
[1461 rows x 2 columns]
Metrolinx_Lakeshore East_UNION STATION_Data Prophet MSE Error: 3449942.1
Metrolinx_Lakeshore East_UNION STATION_Data Prophet RMSE Error: 1857.401976
Metrolinx_Lakeshore East_UNION STATION_Data Prophet R square: 0.4935463283
=================Metrolinx_Lakeshore East_OSHAWA GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Lakeshore East_OSHAWA GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-02 2623 36
2019-01-03 11969 56
2019-01-04 13387 57
2019-01-05 12218 52
2019-01-06 3981 35
Data:
ridership
date
2019-01-02 2623
2019-01-03 11969
2019-01-04 13387
2019-01-05 12218
2019-01-06 3981
Data Regressors:
trips
date
2019-01-02 36
2019-01-03 56
2019-01-04 57
2019-01-05 52
2019-01-06 35
ridership
date
2019-01-02 2623
2019-01-03 11969
2019-01-04 13387
2019-01-05 12218
2019-01-06 3981
... ...
2022-10-02 8176
2022-10-03 6295
2022-10-04 8021
2022-10-05 9921
2022-10-06 10864
[1374 rows x 1 columns]
--------------
Test statistic: -1.6803473394652717
p-value: 0.4412654998346133
Critical Values: {'1%': -3.4351995694971165, '5%': -2.8636817142206836, '10%': -2.5679102510737173}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=19547.848, Time=0.11 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=19645.144, Time=0.01 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=19635.048, Time=0.11 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=19264.360, Time=0.59 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=19643.144, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=19644.116, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=19265.850, Time=0.94 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=19267.324, Time=1.89 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=19399.414, Time=0.51 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=19174.718, Time=2.27 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=19115.700, Time=2.27 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=19067.399, Time=1.56 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=19187.424, Time=1.23 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=19469.376, Time=0.76 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=19181.155, Time=1.71 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=19025.330, Time=1.14 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=19124.409, Time=0.39 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=19094.833, Time=0.82 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=19017.851, Time=1.56 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=19058.053, Time=1.33 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=19040.937, Time=2.98 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=19115.113, Time=2.83 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=19016.201, Time=1.53 sec ARIMA(0,1,2)(1,0,2)[6] : AIC=19056.826, Time=0.56 sec ARIMA(0,1,2)(2,0,1)[6] : AIC=19024.003, Time=0.61 sec ARIMA(0,1,2)(1,0,1)[6] : AIC=19122.559, Time=0.35 sec ARIMA(0,1,1)(2,0,2)[6] : AIC=19114.387, Time=1.95 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=19038.675, Time=2.66 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=19113.066, Time=2.17 sec Best model: ARIMA(0,1,2)(2,0,2)[6] Total fit time: 34.925 seconds Elapsed time: 34.96697680000034
Metrolinx_Lakeshore East_OSHAWA GO_Data ARIMA MSE Error: 10600199.99
Metrolinx_Lakeshore East_OSHAWA GO_Data ARIMA RMSE Error: 3255.794832
Metrolinx_Lakeshore East_OSHAWA GO_Data ARIMA R square: -0.2612499254
Model: "sequential_7"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_7 (LSTM) (None, 100) 40800
dense_7 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
979/979 [==============================] - 4s 4ms/step - loss: 0.0301 Epoch 2/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0253 Epoch 3/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0238 Epoch 4/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0234 Epoch 5/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0177 Epoch 6/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0142 Epoch 7/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0101 Epoch 8/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0115 Epoch 9/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0149 Epoch 10/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0111 Epoch 11/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0095 Epoch 12/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0079 Epoch 13/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0073 Epoch 14/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0069 Epoch 15/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0068 Epoch 16/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 17/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0065 Epoch 18/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 19/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 20/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 21/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 22/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 23/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 24/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 25/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 26/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 27/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 28/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 29/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 30/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 31/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 32/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 33/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 34/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 35/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 36/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 37/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 38/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 39/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0099 Epoch 40/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 41/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 42/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 43/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 44/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 45/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 46/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 47/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 48/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 49/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 50/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 51/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 52/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 53/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 54/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 55/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 56/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 57/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 58/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 59/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 60/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 61/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 62/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 63/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 64/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 65/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 66/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 67/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 68/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 69/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 70/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 71/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0271 Epoch 72/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 73/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 74/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 75/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 76/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 77/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 78/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 79/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 80/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 81/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 82/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 83/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 84/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 85/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 86/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 87/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 88/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 89/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 90/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 91/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 92/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 93/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 94/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 95/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 96/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 97/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 98/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 99/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 100/100 979/979 [==============================] - 4s 4ms/step - loss: 0.0032 Elapsed time: 396.8048272999995
Metrolinx_Lakeshore East_OSHAWA GO_Data LSTM MSE Error: 46493415.28
Metrolinx_Lakeshore East_OSHAWA GO_Data LSTM RMSE Error: 6818.608016
Metrolinx_Lakeshore East_OSHAWA GO_Data LSTM R square: -4.531953794
ds y trips
0 2019-01-02 2623 36
1 2019-01-03 11969 56
2 2019-01-04 13387 57
3 2019-01-05 12218 52
4 2019-01-06 3981 35
########## Cross validation for Metrolinx_Lakeshore East_OSHAWA GO_Data #########
17:16:03 - cmdstanpy - INFO - Chain [1] start processing 17:16:04 - cmdstanpy - INFO - Chain [1] done processing
17:16:05 - cmdstanpy - INFO - Chain [1] start processing 17:16:05 - cmdstanpy - INFO - Chain [1] done processing 17:16:06 - cmdstanpy - INFO - Chain [1] start processing 17:16:07 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 4.521884699999646
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Lakeshore East_OSHAWA GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 39587073.985383
rmse 6054.992311
mae 5315.182371
mdape 1.110105
smape 1.274817
coverage 0.277866
dtype: object
ds trips
0 2019-01-02 36
1 2019-01-03 56
2 2019-01-04 57
3 2019-01-05 52
4 2019-01-06 35
... ... ...
1369 2022-10-02 37
1370 2022-10-03 37
1371 2022-10-04 41
1372 2022-10-05 41
1373 2022-10-06 41
[1374 rows x 2 columns]
Metrolinx_Lakeshore East_OSHAWA GO_Data Prophet MSE Error: 3716981.075
Metrolinx_Lakeshore East_OSHAWA GO_Data Prophet RMSE Error: 1927.947374
Metrolinx_Lakeshore East_OSHAWA GO_Data Prophet R square: 0.5577402211
=================Metrolinx_Lakeshore West_UNION STATION_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Lakeshore West_UNION STATION_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-02 14 1
2019-01-03 3742 10
2019-01-04 3874 10
2019-01-05 3410 10
2019-01-06 21 1
Data:
ridership
date
2019-01-02 14
2019-01-03 3742
2019-01-04 3874
2019-01-05 3410
2019-01-06 21
Data Regressors:
trips
date
2019-01-02 1
2019-01-03 10
2019-01-04 10
2019-01-05 10
2019-01-06 1
ridership
date
2019-01-02 14
2019-01-03 3742
2019-01-04 3874
2019-01-05 3410
2019-01-06 21
... ...
2022-12-28 4528
2022-12-29 5571
2022-12-30 6121
2022-12-31 6325
2023-01-01 3538
[1461 rows x 1 columns]
--------------
Test statistic: -1.8741521512186194
p-value: 0.34431672307352146
Critical Values: {'1%': -3.4349024693573584, '5%': -2.8635506057382325, '10%': -2.5678404322793846}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=21476.977, Time=0.13 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=21571.104, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=21559.010, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=21246.325, Time=0.58 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=21569.105, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=21571.036, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=21247.507, Time=0.86 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=21248.985, Time=1.99 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=21390.660, Time=0.22 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=21149.440, Time=2.12 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=20876.027, Time=2.68 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=21152.405, Time=0.65 sec ARIMA(0,1,0)(2,0,2)[6] intercept : AIC=21371.690, Time=0.93 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=21050.090, Time=2.37 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=20821.078, Time=2.36 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=21004.271, Time=1.42 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=20969.191, Time=0.60 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=21066.035, Time=0.39 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=20997.066, Time=3.04 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=20818.999, Time=2.23 sec ARIMA(0,1,2)(1,0,2)[6] : AIC=21002.277, Time=0.63 sec ARIMA(0,1,2)(2,0,1)[6] : AIC=20966.937, Time=0.53 sec ARIMA(0,1,2)(1,0,1)[6] : AIC=21064.016, Time=0.33 sec ARIMA(0,1,1)(2,0,2)[6] : AIC=20866.260, Time=2.24 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=20995.019, Time=2.49 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=21048.088, Time=1.94 sec Best model: ARIMA(0,1,2)(2,0,2)[6] Total fit time: 30.941 seconds Elapsed time: 30.981863099999828
Metrolinx_Lakeshore West_UNION STATION_Data ARIMA MSE Error: 43211666.99
Metrolinx_Lakeshore West_UNION STATION_Data ARIMA RMSE Error: 6573.558168
Metrolinx_Lakeshore West_UNION STATION_Data ARIMA R square: -1.646225134
Model: "sequential_8"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_8 (LSTM) (None, 100) 40800
dense_8 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1066/1066 [==============================] - 5s 4ms/step - loss: 0.0217 Epoch 2/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0181 Epoch 3/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0172 Epoch 4/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0145 Epoch 5/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0113 Epoch 6/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0086 Epoch 7/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 8/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0055 Epoch 9/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0056 Epoch 10/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 11/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 12/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 13/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0056 Epoch 14/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0056 Epoch 15/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 16/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0047 Epoch 17/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 18/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 19/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0045 Epoch 20/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 21/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 22/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 23/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 24/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 25/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 26/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 27/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 28/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 29/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 30/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 31/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 32/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 33/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 34/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 35/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 36/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 37/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 38/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 39/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 40/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 41/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 42/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 43/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 44/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 45/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 46/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 47/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 48/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 49/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 50/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 51/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 52/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 53/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0034 Epoch 54/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0036 Epoch 55/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0032 Epoch 56/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0033 Epoch 57/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0035 Epoch 58/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0032 Epoch 59/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0033 Epoch 60/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0033 Epoch 61/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 62/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 63/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0030 Epoch 64/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0030 Epoch 65/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 66/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 67/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 68/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0029 Epoch 69/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 70/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 71/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 72/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 73/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 74/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 75/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 76/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 77/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 78/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 79/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 80/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 81/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 82/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 83/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 84/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 85/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 86/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 87/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 88/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 89/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 90/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 91/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 92/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 93/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 94/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 95/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 96/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 97/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 98/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 99/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 100/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Elapsed time: 430.96347200000037
Metrolinx_Lakeshore West_UNION STATION_Data LSTM MSE Error: 48525165.44
Metrolinx_Lakeshore West_UNION STATION_Data LSTM RMSE Error: 6966.000678
Metrolinx_Lakeshore West_UNION STATION_Data LSTM R square: -1.971616727
ds y trips
0 2019-01-02 14 1
1 2019-01-03 3742 10
2 2019-01-04 3874 10
3 2019-01-05 3410 10
4 2019-01-06 21 1
########## Cross validation for Metrolinx_Lakeshore West_UNION STATION_Data #########
17:24:05 - cmdstanpy - INFO - Chain [1] start processing 17:24:05 - cmdstanpy - INFO - Chain [1] done processing
17:24:06 - cmdstanpy - INFO - Chain [1] start processing 17:24:06 - cmdstanpy - INFO - Chain [1] done processing 17:24:07 - cmdstanpy - INFO - Chain [1] start processing 17:24:07 - cmdstanpy - INFO - Chain [1] done processing 17:24:08 - cmdstanpy - INFO - Chain [1] start processing 17:24:08 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.123114099999839
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Lakeshore West_UNION STATION_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 83539219.605652
rmse 8986.225595
mae 7728.97817
mdape 1.389626
smape 1.505438
coverage 0.228977
dtype: object
ds trips
0 2019-01-02 1
1 2019-01-03 10
2 2019-01-04 10
3 2019-01-05 10
4 2019-01-06 1
... ... ...
1456 2022-12-28 39
1457 2022-12-29 46
1458 2022-12-30 45
1459 2022-12-31 45
1460 2023-01-01 40
[1461 rows x 2 columns]
Metrolinx_Lakeshore West_UNION STATION_Data Prophet MSE Error: 8146776.444
Metrolinx_Lakeshore West_UNION STATION_Data Prophet RMSE Error: 2854.255848
Metrolinx_Lakeshore West_UNION STATION_Data Prophet R square: 0.5011022232
=================Metrolinx_Lakeshore West_ALDERSHOT GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Lakeshore West_ALDERSHOT GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-02 3321 35
2019-01-03 7745 34
2019-01-04 9647 34
2019-01-05 8787 33
2019-01-06 5827 33
Data:
ridership
date
2019-01-02 3321
2019-01-03 7745
2019-01-04 9647
2019-01-05 8787
2019-01-06 5827
Data Regressors:
trips
date
2019-01-02 35
2019-01-03 34
2019-01-04 34
2019-01-05 33
2019-01-06 33
ridership
date
2019-01-02 3321
2019-01-03 7745
2019-01-04 9647
2019-01-05 8787
2019-01-06 5827
... ...
2022-12-28 2784
2022-12-29 2798
2022-12-30 3389
2022-12-31 3333
2023-01-01 2324
[1461 rows x 1 columns]
--------------
Test statistic: -1.9055353332243
p-value: 0.329461612773491
Critical Values: {'1%': -3.4349024693573584, '5%': -2.8635506057382325, '10%': -2.5678404322793846}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=19503.272, Time=0.28 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=19516.585, Time=0.01 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=19462.951, Time=0.10 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=19239.773, Time=0.49 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=19514.588, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=19272.022, Time=0.27 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=19234.736, Time=0.58 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=19250.285, Time=0.51 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=19161.535, Time=1.64 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=19211.475, Time=1.15 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=19098.727, Time=2.82 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=19184.885, Time=2.17 sec ARIMA(0,1,0)(2,0,2)[6] intercept : AIC=inf, Time=1.17 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=inf, Time=2.32 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=inf, Time=3.09 sec ARIMA(1,1,0)(2,0,2)[6] intercept : AIC=inf, Time=1.84 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=inf, Time=3.61 sec ARIMA(0,1,1)(2,0,2)[6] : AIC=19094.247, Time=2.08 sec ARIMA(0,1,1)(1,0,2)[6] : AIC=19183.489, Time=1.09 sec ARIMA(0,1,1)(2,0,1)[6] : AIC=19160.166, Time=0.83 sec ARIMA(0,1,1)(1,0,1)[6] : AIC=19232.158, Time=0.47 sec ARIMA(0,1,0)(2,0,2)[6] : AIC=inf, Time=1.00 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=inf, Time=2.08 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=19134.391, Time=1.37 sec ARIMA(1,1,0)(2,0,2)[6] : AIC=inf, Time=1.36 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=inf, Time=2.37 sec Best model: ARIMA(0,1,1)(2,0,2)[6] Total fit time: 34.727 seconds Elapsed time: 34.76149780000014
Metrolinx_Lakeshore West_ALDERSHOT GO_Data ARIMA MSE Error: 12651263.94
Metrolinx_Lakeshore West_ALDERSHOT GO_Data ARIMA RMSE Error: 3556.86153
Metrolinx_Lakeshore West_ALDERSHOT GO_Data ARIMA R square: -1.472951517
Model: "sequential_9"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_9 (LSTM) (None, 100) 40800
dense_9 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1066/1066 [==============================] - 5s 4ms/step - loss: 0.0050 Epoch 2/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 3/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 4/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0043 Epoch 5/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 6/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 7/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 8/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 9/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 10/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 11/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 12/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 13/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 14/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 15/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 16/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0034 Epoch 17/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 18/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 19/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0033 Epoch 20/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0032 Epoch 21/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 22/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0031 Epoch 23/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0030 Epoch 24/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 25/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0030 Epoch 26/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 27/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0029 Epoch 28/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0029 Epoch 29/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 30/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0027 Epoch 31/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0026 Epoch 32/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0028 Epoch 33/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0027 Epoch 34/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 35/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0026 Epoch 36/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0026 Epoch 37/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0025 Epoch 38/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0025 Epoch 39/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0025 Epoch 40/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0024 Epoch 41/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0024 Epoch 42/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0022 Epoch 43/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0024 Epoch 44/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0023 Epoch 45/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0023 Epoch 46/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0022 Epoch 47/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0021 Epoch 48/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0023 Epoch 49/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0022 Epoch 50/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0023 Epoch 51/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0020 Epoch 52/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0021 Epoch 53/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0020 Epoch 54/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0020 Epoch 55/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0020 Epoch 56/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0020 Epoch 57/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0021 Epoch 58/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0020 Epoch 59/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0020 Epoch 60/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0019 Epoch 61/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0020 Epoch 62/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0020 Epoch 63/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0020 Epoch 64/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0019 Epoch 65/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0020 Epoch 66/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0019 Epoch 67/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0019 Epoch 68/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0019 Epoch 69/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0018 Epoch 70/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0019 Epoch 71/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 72/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0019 Epoch 73/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 74/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 75/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 76/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0018 Epoch 77/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 78/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 79/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0018 Epoch 80/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 81/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 82/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 83/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0016 Epoch 84/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0020 Epoch 85/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 86/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0017 Epoch 87/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 88/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0018 Epoch 89/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 90/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0016 Epoch 91/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0017 Epoch 92/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 93/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0016 Epoch 94/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 95/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 96/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0019 Epoch 97/100 1066/1066 [==============================] - 5s 5ms/step - loss: 0.0016 Epoch 98/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 99/100 1066/1066 [==============================] - 4s 4ms/step - loss: 0.0017 Epoch 100/100 1066/1066 [==============================] - 5s 4ms/step - loss: 0.0017 Elapsed time: 450.4398621999999
Metrolinx_Lakeshore West_ALDERSHOT GO_Data LSTM MSE Error: 13807145.21
Metrolinx_Lakeshore West_ALDERSHOT GO_Data LSTM RMSE Error: 3715.796713
Metrolinx_Lakeshore West_ALDERSHOT GO_Data LSTM R square: -1.698892447
ds y trips
0 2019-01-02 3321 35
1 2019-01-03 7745 34
2 2019-01-04 9647 34
3 2019-01-05 8787 33
4 2019-01-06 5827 33
########## Cross validation for Metrolinx_Lakeshore West_ALDERSHOT GO_Data #########
17:32:29 - cmdstanpy - INFO - Chain [1] start processing 17:32:30 - cmdstanpy - INFO - Chain [1] done processing
17:32:31 - cmdstanpy - INFO - Chain [1] start processing 17:32:31 - cmdstanpy - INFO - Chain [1] done processing 17:32:32 - cmdstanpy - INFO - Chain [1] start processing 17:32:32 - cmdstanpy - INFO - Chain [1] done processing 17:32:33 - cmdstanpy - INFO - Chain [1] start processing 17:32:34 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.650375100000019
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Lakeshore West_ALDERSHOT GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 18930892.298948
rmse 4175.067308
mae 3819.049693
mdape 1.448795
smape 1.735109
coverage 0.172927
dtype: object
ds trips
0 2019-01-02 35
1 2019-01-03 34
2 2019-01-04 34
3 2019-01-05 33
4 2019-01-06 33
... ... ...
1456 2022-12-28 18
1457 2022-12-29 16
1458 2022-12-30 16
1459 2022-12-31 16
1460 2023-01-01 18
[1461 rows x 2 columns]
Metrolinx_Lakeshore West_ALDERSHOT GO_Data Prophet MSE Error: 1391344.821
Metrolinx_Lakeshore West_ALDERSHOT GO_Data Prophet RMSE Error: 1179.552805
Metrolinx_Lakeshore West_ALDERSHOT GO_Data Prophet R square: 0.7280328432
=================Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 2751 4
2019-01-04 2786 4
2019-01-05 2469 4
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 2751
2019-01-04 2786
2019-01-05 2469
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 4
2019-01-04 4
2019-01-05 4
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 2751
2019-01-04 2786
2019-01-05 2469
2019-01-06 0
2019-01-07 0
... ...
2022-12-27 0
2022-12-28 0
2022-12-29 247
2022-12-30 340
2022-12-31 245
[1459 rows x 1 columns]
--------------
Test statistic: -1.9075704502786883
p-value: 0.3285081559970963
Critical Values: {'1%': -3.434911997169608, '5%': -2.863554810504947, '10%': -2.567842671398422}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=18759.912, Time=0.35 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=18881.646, Time=0.01 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=18862.133, Time=0.11 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=18427.008, Time=0.48 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=18879.650, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=18876.497, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=18429.008, Time=0.85 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=18429.008, Time=1.98 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=18591.104, Time=0.53 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=18315.027, Time=2.38 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=18250.236, Time=3.29 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=18160.973, Time=1.66 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=18314.737, Time=1.47 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=18631.598, Time=0.94 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=18158.279, Time=2.01 sec ARIMA(1,1,1)(1,0,1)[6] intercept : AIC=18354.387, Time=1.06 sec ARIMA(1,1,1)(2,0,0)[6] intercept : AIC=18287.620, Time=1.92 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=18257.170, Time=2.71 sec ARIMA(1,1,1)(1,0,0)[6] intercept : AIC=18477.276, Time=0.83 sec ARIMA(1,1,1)(1,0,2)[6] intercept : AIC=18238.708, Time=2.36 sec ARIMA(1,1,0)(2,0,1)[6] intercept : AIC=18582.855, Time=1.30 sec ARIMA(2,1,1)(2,0,1)[6] intercept : AIC=18136.368, Time=2.99 sec ARIMA(2,1,1)(1,0,1)[6] intercept : AIC=18283.437, Time=1.25 sec ARIMA(2,1,1)(2,0,0)[6] intercept : AIC=18178.253, Time=2.70 sec ARIMA(2,1,1)(2,0,2)[6] intercept : AIC=18042.549, Time=3.34 sec ARIMA(2,1,1)(1,0,2)[6] intercept : AIC=18141.744, Time=2.89 sec ARIMA(2,1,0)(2,0,2)[6] intercept : AIC=18020.017, Time=2.69 sec ARIMA(2,1,0)(1,0,2)[6] intercept : AIC=18496.830, Time=1.84 sec ARIMA(2,1,0)(2,0,1)[6] intercept : AIC=18237.162, Time=1.95 sec ARIMA(2,1,0)(1,0,1)[6] intercept : AIC=18615.920, Time=0.65 sec ARIMA(1,1,0)(2,0,2)[6] intercept : AIC=18563.337, Time=1.72 sec ARIMA(2,1,0)(2,0,2)[6] : AIC=18086.350, Time=2.22 sec Best model: ARIMA(2,1,0)(2,0,2)[6] intercept Total fit time: 50.564 seconds Elapsed time: 50.60824019999927
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data ARIMA MSE Error: 1378594.423
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data ARIMA RMSE Error: 1174.135607
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data ARIMA R square: -1.394277378
Model: "sequential_10"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_10 (LSTM) (None, 100) 40800
dense_10 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1064/1064 [==============================] - 5s 4ms/step - loss: 0.0548 Epoch 2/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0422 Epoch 3/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0284 Epoch 4/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0216 Epoch 5/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0180 Epoch 6/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0134 Epoch 7/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0114 Epoch 8/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0106 Epoch 9/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0106 Epoch 10/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0096 Epoch 11/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0094 Epoch 12/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0093 Epoch 13/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0089 Epoch 14/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0081 Epoch 15/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0083 Epoch 16/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0082 Epoch 17/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0085 Epoch 18/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0083 Epoch 19/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0076 Epoch 20/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0079 Epoch 21/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0077 Epoch 22/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0077 Epoch 23/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0077 Epoch 24/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0073 Epoch 25/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0070 Epoch 26/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0079 Epoch 27/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0070 Epoch 28/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0075 Epoch 29/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0065 Epoch 30/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0071 Epoch 31/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0075 Epoch 32/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 33/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0075 Epoch 34/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0070 Epoch 35/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0070 Epoch 36/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0068 Epoch 37/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0067 Epoch 38/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0075 Epoch 39/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0070 Epoch 40/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0068 Epoch 41/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 42/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0073 Epoch 43/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0063 Epoch 44/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 45/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 46/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 47/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0067 Epoch 48/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0063 Epoch 49/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 50/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0069 Epoch 51/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0065 Epoch 52/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 53/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0063 Epoch 54/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 55/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0062 Epoch 56/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 57/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 58/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0062 Epoch 59/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 60/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 61/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0060 Epoch 62/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0063 Epoch 63/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0058 Epoch 64/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0068 Epoch 65/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 66/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 67/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 68/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0069 Epoch 69/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0063 Epoch 70/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0057 Epoch 71/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0055 Epoch 72/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 73/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 74/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 75/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0060 Epoch 76/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0054 Epoch 77/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 78/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0059 Epoch 79/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 80/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 81/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 82/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 83/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 84/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0051 Epoch 85/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0054 Epoch 86/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 87/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 88/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 89/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0050 Epoch 90/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 91/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0054 Epoch 92/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 93/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0053 Epoch 94/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 95/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0050 Epoch 96/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0046 Epoch 97/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0051 Epoch 98/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0052 Epoch 99/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0049 Epoch 100/100 1064/1064 [==============================] - 4s 4ms/step - loss: 0.0051 Elapsed time: 432.89158900000075
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data LSTM MSE Error: 1088855.131
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data LSTM RMSE Error: 1043.482214
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data LSTM R square: -0.8910719241
ds y trips
0 2019-01-03 2751 4
1 2019-01-04 2786 4
2 2019-01-05 2469 4
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data #########
17:40:54 - cmdstanpy - INFO - Chain [1] start processing 17:40:54 - cmdstanpy - INFO - Chain [1] done processing
17:40:54 - cmdstanpy - INFO - Chain [1] start processing 17:40:55 - cmdstanpy - INFO - Chain [1] done processing 17:40:56 - cmdstanpy - INFO - Chain [1] start processing 17:40:56 - cmdstanpy - INFO - Chain [1] done processing 17:40:57 - cmdstanpy - INFO - Chain [1] start processing 17:40:57 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.197333500000241
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 2111286.834597
rmse 1439.409986
mae 1270.639686
mdape 2.505621
smape 1.566958
coverage 0.243877
dtype: object
ds trips
0 2019-01-03 4
1 2019-01-04 4
2 2019-01-05 4
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1454 2022-12-27 0
1455 2022-12-28 0
1456 2022-12-29 4
1457 2022-12-30 4
1458 2022-12-31 4
[1459 rows x 2 columns]
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data Prophet MSE Error: 309583.2597
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data Prophet RMSE Error: 556.4020666
Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data Prophet R square: 0.4623304849
=================Metrolinx_Kitchener_UNION STATION_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Kitchener_UNION STATION_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 6020 14
2019-01-04 6412 14
2019-01-05 5651 14
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 6020
2019-01-04 6412
2019-01-05 5651
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 14
2019-01-04 14
2019-01-05 14
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 6020
2019-01-04 6412
2019-01-05 5651
2019-01-06 0
2019-01-07 0
... ...
2022-12-28 0
2022-12-29 1888
2022-12-30 1980
2022-12-31 1789
2023-01-01 200
[1460 rows x 1 columns]
--------------
Test statistic: -2.071415744673781
p-value: 0.2561626879461736
Critical Values: {'1%': -3.434908816804013, '5%': -2.863553406963303, '10%': -2.5678419239852994}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=20106.896, Time=0.35 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=20257.064, Time=0.01 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=20241.893, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=19747.143, Time=0.68 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=20255.068, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=20246.129, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=19748.683, Time=0.99 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=19748.292, Time=2.36 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=19954.378, Time=0.60 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=19627.226, Time=2.65 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=19638.777, Time=2.95 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=19511.361, Time=1.81 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=19653.013, Time=1.46 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=20020.467, Time=0.79 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=19553.732, Time=1.96 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=19486.803, Time=1.14 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=19613.375, Time=1.08 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=19575.865, Time=0.92 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=19485.799, Time=2.14 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=19523.290, Time=1.39 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=19557.733, Time=3.01 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=19697.814, Time=2.95 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=19484.282, Time=1.72 sec ARIMA(0,1,2)(1,0,2)[6] : AIC=19522.782, Time=0.56 sec ARIMA(0,1,2)(2,0,1)[6] : AIC=19486.062, Time=0.50 sec ARIMA(0,1,2)(1,0,1)[6] : AIC=19613.758, Time=0.33 sec ARIMA(0,1,1)(2,0,2)[6] : AIC=19408.934, Time=2.54 sec ARIMA(0,1,1)(1,0,2)[6] : AIC=19626.586, Time=1.39 sec ARIMA(0,1,1)(2,0,1)[6] : AIC=19503.495, Time=1.58 sec ARIMA(0,1,1)(1,0,1)[6] : AIC=19747.628, Time=0.57 sec ARIMA(0,1,0)(2,0,2)[6] : AIC=19826.304, Time=0.95 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=19438.207, Time=2.56 sec ARIMA(1,1,0)(2,0,2)[6] : AIC=19886.144, Time=1.44 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=19521.009, Time=3.18 sec Best model: ARIMA(0,1,1)(2,0,2)[6] Total fit time: 46.758 seconds Elapsed time: 46.80269070000031
Metrolinx_Kitchener_UNION STATION_Data ARIMA MSE Error: 3597639.754
Metrolinx_Kitchener_UNION STATION_Data ARIMA RMSE Error: 1896.744515
Metrolinx_Kitchener_UNION STATION_Data ARIMA R square: -0.2473888758
Model: "sequential_11"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_11 (LSTM) (None, 100) 40800
dense_11 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1065/1065 [==============================] - 5s 4ms/step - loss: 0.0543 Epoch 2/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0483 Epoch 3/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0440 Epoch 4/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0320 Epoch 5/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0190 Epoch 6/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0137 Epoch 7/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0108 Epoch 8/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0097 Epoch 9/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0084 Epoch 10/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0081 Epoch 11/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0084 Epoch 12/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0080 Epoch 13/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0076 Epoch 14/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0075 Epoch 15/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0070 Epoch 16/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0067 Epoch 17/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0068 Epoch 18/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0067 Epoch 19/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0067 Epoch 20/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0066 Epoch 21/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 22/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0067 Epoch 23/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 24/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0061 Epoch 25/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0064 Epoch 26/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0061 Epoch 27/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0060 Epoch 28/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0065 Epoch 29/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0054 Epoch 30/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0057 Epoch 31/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0054 Epoch 32/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0054 Epoch 33/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0055 Epoch 34/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0054 Epoch 35/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0054 Epoch 36/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 37/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 38/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0050 Epoch 39/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0049 Epoch 40/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 41/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0048 Epoch 42/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0048 Epoch 43/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 44/100 1065/1065 [==============================] - 4s 4ms/step - loss: 16527.7305 Epoch 45/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0043 Epoch 46/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 47/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 48/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 49/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 50/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 51/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 52/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0036 Epoch 53/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0035 Epoch 54/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 55/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0033 Epoch 56/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0036 Epoch 57/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0036 Epoch 58/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 59/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0034 Epoch 60/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 61/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 62/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 63/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0035 Epoch 64/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 65/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 66/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 67/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0033 Epoch 68/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 69/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 70/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0035 Epoch 71/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 72/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0034 Epoch 73/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0036 Epoch 74/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0037 Epoch 75/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 76/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 77/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0037 Epoch 78/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 79/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 80/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0042 Epoch 81/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0039 Epoch 82/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 83/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0044 Epoch 84/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0045 Epoch 85/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0045 Epoch 86/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 87/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 88/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 89/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0039 Epoch 90/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 91/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0043 Epoch 92/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0037 Epoch 93/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0042 Epoch 94/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0040 Epoch 95/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0041 Epoch 96/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0041 Epoch 97/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Epoch 98/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0037 Epoch 99/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0038 Epoch 100/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0035 Elapsed time: 446.77511400000003
Metrolinx_Kitchener_UNION STATION_Data LSTM MSE Error: 6772796.138
Metrolinx_Kitchener_UNION STATION_Data LSTM RMSE Error: 2602.459632
Metrolinx_Kitchener_UNION STATION_Data LSTM R square: -1.34829253
ds y trips
0 2019-01-03 6020 14
1 2019-01-04 6412 14
2 2019-01-05 5651 14
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Kitchener_UNION STATION_Data #########
17:49:27 - cmdstanpy - INFO - Chain [1] start processing 17:49:27 - cmdstanpy - INFO - Chain [1] done processing
17:49:28 - cmdstanpy - INFO - Chain [1] start processing 17:49:28 - cmdstanpy - INFO - Chain [1] done processing 17:49:29 - cmdstanpy - INFO - Chain [1] start processing 17:49:30 - cmdstanpy - INFO - Chain [1] done processing 17:49:31 - cmdstanpy - INFO - Chain [1] start processing 17:49:31 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.415826200000083
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Kitchener_UNION STATION_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 13822633.23599
rmse 3645.949825
mae 3210.018857
mdape 1.797264
smape 1.667121
coverage 0.338827
dtype: object
ds trips
0 2019-01-03 14
1 2019-01-04 14
2 2019-01-05 14
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1455 2022-12-28 0
1456 2022-12-29 24
1457 2022-12-30 24
1458 2022-12-31 24
1459 2023-01-01 2
[1460 rows x 2 columns]
Metrolinx_Kitchener_UNION STATION_Data Prophet MSE Error: 861284.5365
Metrolinx_Kitchener_UNION STATION_Data Prophet RMSE Error: 928.0541668
Metrolinx_Kitchener_UNION STATION_Data Prophet R square: 0.7013717817
=================Metrolinx_Kitchener_KITCHENER GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Kitchener_KITCHENER GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 3221 4
2019-01-04 3365 4
2019-01-05 2879 4
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 3221
2019-01-04 3365
2019-01-05 2879
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 4
2019-01-04 4
2019-01-05 4
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 3221
2019-01-04 3365
2019-01-05 2879
2019-01-06 0
2019-01-07 0
... ...
2022-12-27 0
2022-12-28 0
2022-12-29 1338
2022-12-30 1298
2022-12-31 1204
[1459 rows x 1 columns]
--------------
Test statistic: -2.0596400998693563
p-value: 0.2610499839268926
Critical Values: {'1%': -3.434908816804013, '5%': -2.863553406963303, '10%': -2.5678419239852994}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=19223.915, Time=0.31 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=19377.621, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=19354.143, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=18885.785, Time=0.65 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=19375.624, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=19372.284, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=18887.662, Time=0.96 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=18887.482, Time=2.05 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=19077.945, Time=0.56 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=18768.796, Time=1.92 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=18753.961, Time=2.82 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=18669.040, Time=1.66 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=18783.691, Time=1.32 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=19151.148, Time=0.83 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=18718.658, Time=1.92 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=18622.095, Time=1.72 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=18740.895, Time=0.65 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=18697.996, Time=0.89 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=18620.895, Time=2.72 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=18662.509, Time=1.35 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=18645.562, Time=2.98 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=18659.856, Time=3.10 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=18624.537, Time=0.96 sec Best model: ARIMA(0,1,2)(2,0,2)[6] intercept Total fit time: 29.598 seconds Elapsed time: 29.631757799999832
Metrolinx_Kitchener_KITCHENER GO_Data ARIMA MSE Error: 5939522.787
Metrolinx_Kitchener_KITCHENER GO_Data ARIMA RMSE Error: 2437.113618
Metrolinx_Kitchener_KITCHENER GO_Data ARIMA R square: -2.148101187
Model: "sequential_12"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_12 (LSTM) (None, 100) 40800
dense_12 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1064/1064 [==============================] - 6s 5ms/step - loss: 0.0625 Epoch 2/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0521 Epoch 3/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0323 Epoch 4/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0217 Epoch 5/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0186 Epoch 6/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0155 Epoch 7/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0127 Epoch 8/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0124 Epoch 9/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0110 Epoch 10/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0107 Epoch 11/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0109 Epoch 12/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0104 Epoch 13/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0098 Epoch 14/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0101 Epoch 15/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0095 Epoch 16/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0093 Epoch 17/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0097 Epoch 18/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0087 Epoch 19/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0086 Epoch 20/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0089 Epoch 21/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0083 Epoch 22/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0079 Epoch 23/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0091 Epoch 24/100 1064/1064 [==============================] - 6s 5ms/step - loss: 0.0086 Epoch 25/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0082 Epoch 26/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0081 Epoch 27/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0083 Epoch 28/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0082 Epoch 29/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0081 Epoch 30/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0074 Epoch 31/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0076 Epoch 32/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0074 Epoch 33/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0074 Epoch 34/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0076 Epoch 35/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0073 Epoch 36/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0069 Epoch 37/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0073 Epoch 38/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0069 Epoch 39/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0067 Epoch 40/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0067 Epoch 41/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0069 Epoch 42/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 43/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0067 Epoch 44/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0073 Epoch 45/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0069 Epoch 46/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 47/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 48/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 49/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 50/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 51/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 52/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 53/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 54/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 55/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 56/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 57/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0067 Epoch 58/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 59/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 60/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 61/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 62/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0056 Epoch 63/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 64/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 65/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0055 Epoch 66/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 67/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 68/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0056 Epoch 69/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 70/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 71/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 72/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 73/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 74/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 75/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 76/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 77/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 78/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 79/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 80/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 81/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 82/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 83/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 84/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 85/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 86/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 87/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 88/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 89/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 90/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 91/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 92/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 93/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 94/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 95/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 96/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 97/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 98/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 99/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 100/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0043 Elapsed time: 500.31980850000036
Metrolinx_Kitchener_KITCHENER GO_Data LSTM MSE Error: 3341113.245
Metrolinx_Kitchener_KITCHENER GO_Data LSTM RMSE Error: 1827.871233
Metrolinx_Kitchener_KITCHENER GO_Data LSTM R square: -0.7708767104
ds y trips
0 2019-01-03 3221 4
1 2019-01-04 3365 4
2 2019-01-05 2879 4
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Kitchener_KITCHENER GO_Data #########
17:58:37 - cmdstanpy - INFO - Chain [1] start processing 17:58:38 - cmdstanpy - INFO - Chain [1] done processing
17:58:39 - cmdstanpy - INFO - Chain [1] start processing 17:58:39 - cmdstanpy - INFO - Chain [1] done processing 17:58:40 - cmdstanpy - INFO - Chain [1] start processing 17:58:41 - cmdstanpy - INFO - Chain [1] done processing 17:58:42 - cmdstanpy - INFO - Chain [1] start processing 17:58:42 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 6.405299000000014
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Kitchener_KITCHENER GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 4672639.552511
rmse 2115.681621
mae 1859.079386
mdape 2.193989
smape 1.519568
coverage 0.304407
dtype: object
ds trips
0 2019-01-03 4
1 2019-01-04 4
2 2019-01-05 4
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1454 2022-12-27 0
1455 2022-12-28 0
1456 2022-12-29 8
1457 2022-12-30 8
1458 2022-12-31 8
[1459 rows x 2 columns]
Metrolinx_Kitchener_KITCHENER GO_Data Prophet MSE Error: 428634.4902
Metrolinx_Kitchener_KITCHENER GO_Data Prophet RMSE Error: 654.701833
Metrolinx_Kitchener_KITCHENER GO_Data Prophet R square: 0.7728125986
=================Metrolinx_Kitchener_MOUNT PLEASANT GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Kitchener_MOUNT PLEASANT GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 1280 9
2019-01-04 1581 9
2019-01-05 1353 9
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 1280
2019-01-04 1581
2019-01-05 1353
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 9
2019-01-04 9
2019-01-05 9
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 1280
2019-01-04 1581
2019-01-05 1353
2019-01-06 0
2019-01-07 0
... ...
2022-12-28 0
2022-12-29 724
2022-12-30 714
2022-12-31 715
2023-01-01 152
[1460 rows x 1 columns]
--------------
Test statistic: -2.333156962959321
p-value: 0.1614970300713811
Critical Values: {'1%': -3.434908816804013, '5%': -2.863553406963303, '10%': -2.5678419239852994}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=16774.329, Time=0.33 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=16853.361, Time=0.01 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=16844.879, Time=0.28 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=16333.280, Time=0.56 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=16851.365, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=16574.136, Time=0.16 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=16329.150, Time=0.73 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=16460.492, Time=0.50 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=16140.779, Time=1.52 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=16254.763, Time=1.13 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=16339.591, Time=2.92 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=16215.270, Time=1.92 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=16673.260, Time=0.79 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=16141.458, Time=1.83 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=16133.721, Time=1.60 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=16250.200, Time=0.93 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=16202.347, Time=0.75 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=inf, Time=2.71 sec ARIMA(0,1,2)(1,0,0)[6] intercept : AIC=16304.180, Time=0.58 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=16154.913, Time=1.99 sec ARIMA(1,1,2)(2,0,1)[6] intercept : AIC=16064.582, Time=2.07 sec ARIMA(1,1,2)(1,0,1)[6] intercept : AIC=16189.813, Time=1.12 sec ARIMA(1,1,2)(2,0,0)[6] intercept : AIC=16143.647, Time=1.84 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=16112.650, Time=2.86 sec ARIMA(1,1,2)(1,0,0)[6] intercept : AIC=16275.383, Time=0.68 sec ARIMA(1,1,2)(1,0,2)[6] intercept : AIC=16116.150, Time=2.64 sec ARIMA(2,1,2)(2,0,1)[6] intercept : AIC=16038.615, Time=3.24 sec ARIMA(2,1,2)(1,0,1)[6] intercept : AIC=16199.278, Time=1.47 sec ARIMA(2,1,2)(2,0,0)[6] intercept : AIC=16149.473, Time=2.89 sec ARIMA(2,1,2)(2,0,2)[6] intercept : AIC=16082.100, Time=3.49 sec ARIMA(2,1,2)(1,0,0)[6] intercept : AIC=16069.745, Time=1.39 sec ARIMA(2,1,2)(1,0,2)[6] intercept : AIC=16118.042, Time=3.05 sec ARIMA(2,1,1)(2,0,1)[6] intercept : AIC=16031.114, Time=2.87 sec ARIMA(2,1,1)(1,0,1)[6] intercept : AIC=16192.818, Time=1.27 sec ARIMA(2,1,1)(2,0,0)[6] intercept : AIC=16156.444, Time=2.21 sec ARIMA(2,1,1)(2,0,2)[6] intercept : AIC=16022.348, Time=3.30 sec ARIMA(2,1,1)(1,0,2)[6] intercept : AIC=16110.137, Time=2.69 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=16145.166, Time=3.28 sec ARIMA(2,1,0)(2,0,2)[6] intercept : AIC=16022.590, Time=3.23 sec ARIMA(1,1,0)(2,0,2)[6] intercept : AIC=16445.402, Time=1.86 sec ARIMA(2,1,1)(2,0,2)[6] : AIC=16026.332, Time=2.41 sec Best model: ARIMA(2,1,1)(2,0,2)[6] intercept Total fit time: 71.158 seconds Elapsed time: 71.22309179999866
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data ARIMA MSE Error: 416834.7958
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data ARIMA RMSE Error: 645.6274435
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data ARIMA R square: -2.181842534
Model: "sequential_13"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_13 (LSTM) (None, 100) 40800
dense_13 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1065/1065 [==============================] - 5s 4ms/step - loss: 0.0156 Epoch 2/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0136 Epoch 3/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0939 Epoch 4/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0139 Epoch 5/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0132 Epoch 6/100 1065/1065 [==============================] - 4s 4ms/step - loss: 0.0169 Epoch 7/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0133 Epoch 8/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0133 Epoch 9/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0131 Epoch 10/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0127 Epoch 11/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0125 Epoch 12/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0115 Epoch 13/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0113 Epoch 14/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0094 Epoch 15/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0083 Epoch 16/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0077 Epoch 17/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0073 Epoch 18/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0073 Epoch 19/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0070 Epoch 20/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0068 Epoch 21/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 22/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0061 Epoch 23/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0057 Epoch 24/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 25/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 26/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0043 Epoch 27/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0038 Epoch 28/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 29/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 30/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0036 Epoch 31/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 32/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0035 Epoch 33/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0034 Epoch 34/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0034 Epoch 35/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0035 Epoch 36/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0034 Epoch 37/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0036 Epoch 38/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 39/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 40/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 41/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 42/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0031 Epoch 43/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0034 Epoch 44/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0032 Epoch 45/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0031 Epoch 46/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0032 Epoch 47/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0031 Epoch 48/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0031 Epoch 49/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0030 Epoch 50/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0031 Epoch 51/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 52/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0031 Epoch 53/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0029 Epoch 54/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 55/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0029 Epoch 56/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0030 Epoch 57/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 58/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 59/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0029 Epoch 60/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0029 Epoch 61/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 62/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0029 Epoch 63/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0029 Epoch 64/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 65/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0029 Epoch 66/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 67/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 68/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 69/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0027 Epoch 70/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0029 Epoch 71/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 72/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0027 Epoch 73/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 74/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0031 Epoch 75/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 76/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 77/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 78/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 79/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0027 Epoch 80/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0026 Epoch 81/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0026 Epoch 82/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 83/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0026 Epoch 84/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0024 Epoch 85/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 86/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0026 Epoch 87/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0025 Epoch 88/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0025 Epoch 89/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0024 Epoch 90/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0025 Epoch 91/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 92/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0024 Epoch 93/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0024 Epoch 94/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0025 Epoch 95/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0023 Epoch 96/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0024 Epoch 97/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0023 Epoch 98/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0024 Epoch 99/100 1065/1065 [==============================] - 5s 4ms/step - loss: 0.0024 Epoch 100/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0024 Elapsed time: 485.5262039999998
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data LSTM MSE Error: 250116.7373
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data LSTM RMSE Error: 500.1167237
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data LSTM R square: -0.9092265839
ds y trips
0 2019-01-03 1280 9
1 2019-01-04 1581 9
2 2019-01-05 1353 9
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Kitchener_MOUNT PLEASANT GO_Data #########
18:08:15 - cmdstanpy - INFO - Chain [1] start processing 18:08:16 - cmdstanpy - INFO - Chain [1] done processing
18:08:17 - cmdstanpy - INFO - Chain [1] start processing 18:08:17 - cmdstanpy - INFO - Chain [1] done processing 18:08:18 - cmdstanpy - INFO - Chain [1] start processing 18:08:18 - cmdstanpy - INFO - Chain [1] done processing 18:08:19 - cmdstanpy - INFO - Chain [1] start processing 18:08:20 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.644526699999915
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Kitchener_MOUNT PLEASANT GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 386439.658466
rmse 608.391662
mae 519.015282
mdape 1.580368
smape 1.477632
coverage 0.410121
dtype: object
ds trips
0 2019-01-03 9
1 2019-01-04 9
2 2019-01-05 9
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1455 2022-12-28 0
1456 2022-12-29 6
1457 2022-12-30 6
1458 2022-12-31 6
1459 2023-01-01 1
[1460 rows x 2 columns]
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data Prophet MSE Error: 28303.51396
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data Prophet RMSE Error: 168.2364823
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data Prophet R square: 0.7839495995
=================Metrolinx_Milton_UNION STATION_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Milton_UNION STATION_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 7296 10
2019-01-04 7662 10
2019-01-05 6591 10
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 7296
2019-01-04 7662
2019-01-05 6591
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 10
2019-01-04 10
2019-01-05 10
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 7296
2019-01-04 7662
2019-01-05 6591
2019-01-06 0
2019-01-07 0
... ...
2022-12-27 0
2022-12-28 0
2022-12-29 804
2022-12-30 811
2022-12-31 540
[1459 rows x 1 columns]
--------------
Test statistic: -1.9815053773824545
p-value: 0.29474075472443795
Critical Values: {'1%': -3.434911997169608, '5%': -2.863554810504947, '10%': -2.567842671398422}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=20509.031, Time=0.34 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=20645.505, Time=0.01 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=20621.554, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=20179.566, Time=0.59 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=20643.510, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=20625.294, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=20181.291, Time=0.98 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=20180.823, Time=2.16 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=20620.200, Time=0.17 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=20076.691, Time=2.27 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=20020.126, Time=2.63 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=19920.920, Time=1.65 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=20065.708, Time=1.31 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=20419.881, Time=0.92 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=19940.167, Time=1.87 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=19877.618, Time=1.16 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=19998.805, Time=0.75 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=19951.797, Time=0.86 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=19721.743, Time=3.01 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=19921.246, Time=1.60 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=inf, Time=3.46 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=20028.623, Time=3.19 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=19709.245, Time=2.23 sec ARIMA(0,1,2)(1,0,2)[6] : AIC=19921.022, Time=0.55 sec ARIMA(0,1,2)(2,0,1)[6] : AIC=19877.140, Time=0.60 sec ARIMA(0,1,2)(1,0,1)[6] : AIC=19998.372, Time=0.33 sec ARIMA(0,1,1)(2,0,2)[6] : AIC=19994.222, Time=2.08 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=inf, Time=2.89 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=20027.208, Time=2.57 sec Best model: ARIMA(0,1,2)(2,0,2)[6] Total fit time: 40.392 seconds Elapsed time: 40.43140719999974
Metrolinx_Milton_UNION STATION_Data ARIMA MSE Error: 1818460.487
Metrolinx_Milton_UNION STATION_Data ARIMA RMSE Error: 1348.503054
Metrolinx_Milton_UNION STATION_Data ARIMA R square: -0.2056134299
Model: "sequential_14"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_14 (LSTM) (None, 100) 40800
dense_14 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1064/1064 [==============================] - 5s 4ms/step - loss: 0.0727 Epoch 2/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0645 Epoch 3/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0564 Epoch 4/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0344 Epoch 5/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0243 Epoch 6/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0169 Epoch 7/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0157 Epoch 8/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0128 Epoch 9/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0123 Epoch 10/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0129 Epoch 11/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0118 Epoch 12/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0117 Epoch 13/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0105 Epoch 14/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0111 Epoch 15/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0104 Epoch 16/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0099 Epoch 17/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0101 Epoch 18/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0098 Epoch 19/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0093 Epoch 20/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0089 Epoch 21/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0091 Epoch 22/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0085 Epoch 23/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0083 Epoch 24/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0088 Epoch 25/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0084 Epoch 26/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0079 Epoch 27/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0079 Epoch 28/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0077 Epoch 29/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0079 Epoch 30/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0076 Epoch 31/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0079 Epoch 32/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0078 Epoch 33/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0071 Epoch 34/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0074 Epoch 35/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0067 Epoch 36/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0072 Epoch 37/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0071 Epoch 38/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0069 Epoch 39/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0069 Epoch 40/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 41/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 42/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 43/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 44/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 45/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0061 Epoch 46/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 47/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 48/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 49/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 50/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0057 Epoch 51/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 52/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 53/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 54/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 55/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 56/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 57/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 58/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 59/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 60/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 61/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 62/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 63/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 64/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 65/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 66/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 67/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 68/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.1584 Epoch 69/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 70/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 71/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 72/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 73/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 74/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 75/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 76/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 77/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 78/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 79/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 80/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 81/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 82/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 83/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 84/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 85/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 86/100 1064/1064 [==============================] - 6s 5ms/step - loss: 0.0046 Epoch 87/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 88/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 89/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 90/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0051 Epoch 91/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0043 Epoch 92/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 93/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 94/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0042 Epoch 95/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 96/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0042 Epoch 97/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0050 Epoch 98/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 99/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 100/100 1064/1064 [==============================] - 5s 4ms/step - loss: 0.0045 Elapsed time: 490.4276437000008
Metrolinx_Milton_UNION STATION_Data LSTM MSE Error: 3419330.941
Metrolinx_Milton_UNION STATION_Data LSTM RMSE Error: 1849.143299
Metrolinx_Milton_UNION STATION_Data LSTM R square: -1.266967764
ds y trips
0 2019-01-03 7296 10
1 2019-01-04 7662 10
2 2019-01-05 6591 10
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Milton_UNION STATION_Data #########
18:17:27 - cmdstanpy - INFO - Chain [1] start processing 18:17:28 - cmdstanpy - INFO - Chain [1] done processing
18:17:28 - cmdstanpy - INFO - Chain [1] start processing 18:17:28 - cmdstanpy - INFO - Chain [1] done processing 18:17:29 - cmdstanpy - INFO - Chain [1] start processing 18:17:30 - cmdstanpy - INFO - Chain [1] done processing 18:17:31 - cmdstanpy - INFO - Chain [1] start processing 18:17:31 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.298198300000877
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Milton_UNION STATION_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 10999985.351668
rmse 3233.406031
mae 2830.341322
mdape 3.307929
smape 1.659631
coverage 0.261222
dtype: object
ds trips
0 2019-01-03 10
1 2019-01-04 10
2 2019-01-05 10
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1454 2022-12-27 0
1455 2022-12-28 0
1456 2022-12-29 6
1457 2022-12-30 6
1458 2022-12-31 6
[1459 rows x 2 columns]
Metrolinx_Milton_UNION STATION_Data Prophet MSE Error: 878826.8832
Metrolinx_Milton_UNION STATION_Data Prophet RMSE Error: 937.4576701
Metrolinx_Milton_UNION STATION_Data Prophet R square: 0.4173502804
=================Metrolinx_Milton_MILTON GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Milton_MILTON GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 7304 10
2019-01-04 7665 10
2019-01-05 6737 10
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 7304
2019-01-04 7665
2019-01-05 6737
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 10
2019-01-04 10
2019-01-05 10
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 7304
2019-01-04 7665
2019-01-05 6737
2019-01-06 0
2019-01-07 0
... ...
2022-12-28 0
2022-12-29 695
2022-12-30 670
2022-12-31 469
2023-01-01 72
[1460 rows x 1 columns]
--------------
Test statistic: -1.9570185029217027
p-value: 0.30573094828542036
Critical Values: {'1%': -3.434908816804013, '5%': -2.863553406963303, '10%': -2.5678419239852994}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=20584.352, Time=0.14 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=20735.846, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=20710.172, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=20245.648, Time=0.78 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=20733.852, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=20704.428, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=20248.335, Time=1.06 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=20469.932, Time=0.41 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=20702.694, Time=0.17 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=20132.539, Time=2.34 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=20095.540, Time=2.76 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=20010.664, Time=1.69 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=20134.974, Time=1.32 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=20509.342, Time=0.49 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=20072.532, Time=2.03 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=19932.530, Time=1.20 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=20068.936, Time=0.67 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=20022.130, Time=0.92 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=19788.573, Time=3.03 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=19975.839, Time=1.37 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=20034.296, Time=3.21 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=20107.457, Time=3.16 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=19777.464, Time=2.55 sec ARIMA(0,1,2)(1,0,2)[6] : AIC=19975.703, Time=0.61 sec ARIMA(0,1,2)(2,0,1)[6] : AIC=19932.076, Time=0.59 sec ARIMA(0,1,2)(1,0,1)[6] : AIC=20068.398, Time=0.35 sec ARIMA(0,1,1)(2,0,2)[6] : AIC=20101.170, Time=2.34 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=20030.055, Time=2.77 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=inf, Time=2.68 sec Best model: ARIMA(0,1,2)(2,0,2)[6] Total fit time: 38.858 seconds Elapsed time: 38.90018729999974
Metrolinx_Milton_MILTON GO_Data ARIMA MSE Error: 2189601.27
Metrolinx_Milton_MILTON GO_Data ARIMA RMSE Error: 1479.730134
Metrolinx_Milton_MILTON GO_Data ARIMA R square: -0.2503147977
Model: "sequential_15"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_15 (LSTM) (None, 100) 40800
dense_15 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1065/1065 [==============================] - 6s 5ms/step - loss: 0.0576 Epoch 2/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0487 Epoch 3/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0435 Epoch 4/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0267 Epoch 5/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0223 Epoch 6/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0193 Epoch 7/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0168 Epoch 8/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0196 Epoch 9/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0136 Epoch 10/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0114 Epoch 11/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0096 Epoch 12/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0095 Epoch 13/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0089 Epoch 14/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0089 Epoch 15/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0086 Epoch 16/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0086 Epoch 17/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0077 Epoch 18/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0075 Epoch 19/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0075 Epoch 20/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0074 Epoch 21/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0075 Epoch 22/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0075 Epoch 23/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0068 Epoch 24/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0072 Epoch 25/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 26/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 27/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 28/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0068 Epoch 29/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 30/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 31/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 32/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 33/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 34/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0067 Epoch 35/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 36/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0056 Epoch 37/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 38/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0056 Epoch 39/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 40/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 41/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 42/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 43/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 44/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 45/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 46/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 47/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 48/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 49/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 50/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 51/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 52/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 53/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 54/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 55/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 56/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 57/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 58/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 59/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 60/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 61/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 62/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 63/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 64/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 65/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 66/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 67/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 68/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 69/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 70/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 71/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 72/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0268 Epoch 73/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 74/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 75/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 76/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 77/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 78/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 79/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 80/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 81/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 82/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 83/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 84/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 85/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 86/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 87/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 88/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 89/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 90/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 91/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 92/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0045 Epoch 93/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 94/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0035 Epoch 95/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0032 Epoch 96/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 97/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 98/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 99/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 100/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0037 Elapsed time: 525.3197590999989
Metrolinx_Milton_MILTON GO_Data LSTM MSE Error: 2983731.432
Metrolinx_Milton_MILTON GO_Data LSTM RMSE Error: 1727.348092
Metrolinx_Milton_MILTON GO_Data LSTM R square: -0.7037821513
ds y trips
0 2019-01-03 7304 10
1 2019-01-04 7665 10
2 2019-01-05 6737 10
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Milton_MILTON GO_Data #########
18:27:11 - cmdstanpy - INFO - Chain [1] start processing 18:27:12 - cmdstanpy - INFO - Chain [1] done processing
18:27:12 - cmdstanpy - INFO - Chain [1] start processing 18:27:13 - cmdstanpy - INFO - Chain [1] done processing 18:27:14 - cmdstanpy - INFO - Chain [1] start processing 18:27:14 - cmdstanpy - INFO - Chain [1] done processing 18:27:15 - cmdstanpy - INFO - Chain [1] start processing 18:27:16 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.616625799999383
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Milton_MILTON GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 12176781.793722
rmse 3406.622697
mae 2977.250683
mdape 3.338776
smape 1.663943
coverage 0.280463
dtype: object
ds trips
0 2019-01-03 10
1 2019-01-04 10
2 2019-01-05 10
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1455 2022-12-28 0
1456 2022-12-29 6
1457 2022-12-30 6
1458 2022-12-31 6
1459 2023-01-01 2
[1460 rows x 2 columns]
Metrolinx_Milton_MILTON GO_Data Prophet MSE Error: 1040368.943
Metrolinx_Milton_MILTON GO_Data Prophet RMSE Error: 1019.984776
Metrolinx_Milton_MILTON GO_Data Prophet R square: 0.4059244019
=================Metrolinx_Richmond Hill_UNION STATION_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Richmond Hill_UNION STATION_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 2997 7
2019-01-04 3114 7
2019-01-05 2711 7
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 2997
2019-01-04 3114
2019-01-05 2711
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 7
2019-01-04 7
2019-01-05 7
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 2997
2019-01-04 3114
2019-01-05 2711
2019-01-06 0
2019-01-07 0
... ...
2022-12-27 0
2022-12-28 0
2022-12-29 221
2022-12-30 204
2022-12-31 137
[1459 rows x 1 columns]
--------------
Test statistic: -1.8662966167443402
p-value: 0.34807849507625566
Critical Values: {'1%': -3.434911997169608, '5%': -2.863554810504947, '10%': -2.567842671398422}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=18543.385, Time=0.35 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=18700.668, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=18676.389, Time=0.15 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=18213.348, Time=0.63 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=18698.674, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=18678.454, Time=0.08 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=18215.237, Time=1.02 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=18405.102, Time=0.94 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=18674.378, Time=0.17 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=18100.482, Time=1.87 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=18069.825, Time=2.64 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=17938.316, Time=1.79 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=18097.108, Time=1.64 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=18402.209, Time=1.12 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=17937.599, Time=2.00 sec ARIMA(1,1,1)(1,0,1)[6] intercept : AIC=18194.166, Time=1.07 sec ARIMA(1,1,1)(2,0,0)[6] intercept : AIC=18065.360, Time=1.95 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=18068.015, Time=3.07 sec ARIMA(1,1,1)(1,0,0)[6] intercept : AIC=18296.716, Time=0.85 sec ARIMA(1,1,1)(1,0,2)[6] intercept : AIC=18018.101, Time=2.71 sec ARIMA(1,1,0)(2,0,1)[6] intercept : AIC=18364.092, Time=1.47 sec ARIMA(2,1,1)(2,0,1)[6] intercept : AIC=17953.729, Time=2.90 sec ARIMA(1,1,2)(2,0,1)[6] intercept : AIC=17920.653, Time=2.22 sec ARIMA(1,1,2)(1,0,1)[6] intercept : AIC=18065.888, Time=1.32 sec ARIMA(1,1,2)(2,0,0)[6] intercept : AIC=17956.521, Time=1.79 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=18001.908, Time=3.31 sec ARIMA(1,1,2)(1,0,0)[6] intercept : AIC=18110.456, Time=1.18 sec ARIMA(1,1,2)(1,0,2)[6] intercept : AIC=18066.070, Time=2.64 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=17905.463, Time=1.81 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=18049.226, Time=0.77 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=18002.583, Time=0.92 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=17753.406, Time=3.22 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=17958.746, Time=2.10 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=17751.455, Time=1.84 sec ARIMA(0,1,2)(1,0,2)[6] : AIC=17960.656, Time=0.40 sec ARIMA(0,1,2)(2,0,1)[6] : AIC=17905.355, Time=0.66 sec ARIMA(0,1,2)(1,0,1)[6] : AIC=18048.397, Time=0.52 sec ARIMA(0,1,1)(2,0,2)[6] : AIC=inf, Time=2.13 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=17995.985, Time=2.58 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=18072.188, Time=2.56 sec Best model: ARIMA(0,1,2)(2,0,2)[6] Total fit time: 60.410 seconds Elapsed time: 60.47983639999984
Metrolinx_Richmond Hill_UNION STATION_Data ARIMA MSE Error: 207053.4948
Metrolinx_Richmond Hill_UNION STATION_Data ARIMA RMSE Error: 455.0313119
Metrolinx_Richmond Hill_UNION STATION_Data ARIMA R square: -0.2621875697
Model: "sequential_16"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_16 (LSTM) (None, 100) 40800
dense_16 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1064/1064 [==============================] - 6s 5ms/step - loss: 0.0752 Epoch 2/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0594 Epoch 3/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0335 Epoch 4/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0235 Epoch 5/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0178 Epoch 6/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0145 Epoch 7/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0123 Epoch 8/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0122 Epoch 9/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0109 Epoch 10/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0113 Epoch 11/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0108 Epoch 12/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0101 Epoch 13/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0147 Epoch 14/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0102 Epoch 15/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0092 Epoch 16/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0099 Epoch 17/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0094 Epoch 18/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0097 Epoch 19/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0091 Epoch 20/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0095 Epoch 21/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0088 Epoch 22/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0088 Epoch 23/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0085 Epoch 24/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0081 Epoch 25/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0078 Epoch 26/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0087 Epoch 27/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0077 Epoch 28/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0079 Epoch 29/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 30/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0078 Epoch 31/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 32/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 33/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.3415 Epoch 34/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0079 Epoch 35/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 36/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 37/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 38/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 39/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 40/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 41/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 42/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 43/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 44/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 45/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 46/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 47/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 48/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 49/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 50/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 51/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 52/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 53/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 54/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 55/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 56/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 57/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0065 Epoch 58/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0056 Epoch 59/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0057 Epoch 60/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 61/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0056 Epoch 62/100 1064/1064 [==============================] - 6s 5ms/step - loss: 0.0062 Epoch 63/100 1064/1064 [==============================] - 6s 5ms/step - loss: 0.0053 Epoch 64/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 65/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 66/100 1064/1064 [==============================] - 6s 6ms/step - loss: 0.0053 Epoch 67/100 1064/1064 [==============================] - 6s 5ms/step - loss: 0.0049 Epoch 68/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 69/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 70/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 71/100 1064/1064 [==============================] - 6s 6ms/step - loss: 0.0051 Epoch 72/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 73/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 74/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 75/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 76/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 77/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 78/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 79/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 80/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 81/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 82/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 83/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 84/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 85/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 86/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 87/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 88/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 89/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 90/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 91/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 92/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 93/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 94/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 95/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 96/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 97/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 98/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 99/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 100/100 1064/1064 [==============================] - 5s 5ms/step - loss: 0.0040 Elapsed time: 510.1176524000002
Metrolinx_Richmond Hill_UNION STATION_Data LSTM MSE Error: 321904.1676
Metrolinx_Richmond Hill_UNION STATION_Data LSTM RMSE Error: 567.3659908
Metrolinx_Richmond Hill_UNION STATION_Data LSTM R square: -0.9623114274
ds y trips
0 2019-01-03 2997 7
1 2019-01-04 3114 7
2 2019-01-05 2711 7
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Richmond Hill_UNION STATION_Data #########
18:37:03 - cmdstanpy - INFO - Chain [1] start processing 18:37:03 - cmdstanpy - INFO - Chain [1] done processing
18:37:04 - cmdstanpy - INFO - Chain [1] start processing 18:37:04 - cmdstanpy - INFO - Chain [1] done processing 18:37:05 - cmdstanpy - INFO - Chain [1] start processing 18:37:06 - cmdstanpy - INFO - Chain [1] done processing 18:37:07 - cmdstanpy - INFO - Chain [1] start processing 18:37:07 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.313034999999218
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Richmond Hill_UNION STATION_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 1443266.762784
rmse 1178.484107
mae 1041.797992
mdape 3.544262
smape 1.68436
coverage 0.283772
dtype: object
ds trips
0 2019-01-03 7
1 2019-01-04 7
2 2019-01-05 7
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1454 2022-12-27 0
1455 2022-12-28 0
1456 2022-12-29 4
1457 2022-12-30 4
1458 2022-12-31 4
[1459 rows x 2 columns]
Metrolinx_Richmond Hill_UNION STATION_Data Prophet MSE Error: 185533.2098
Metrolinx_Richmond Hill_UNION STATION_Data Prophet RMSE Error: 430.7356612
Metrolinx_Richmond Hill_UNION STATION_Data Prophet R square: -0.1310010074
=================Metrolinx_Stouffville_UNION STATION_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Stouffville_UNION STATION_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 4499 14
2019-01-04 4783 15
2019-01-05 4234 15
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 4499
2019-01-04 4783
2019-01-05 4234
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 14
2019-01-04 15
2019-01-05 15
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 4499
2019-01-04 4783
2019-01-05 4234
2019-01-06 0
2019-01-07 0
... ...
2022-12-28 778
2022-12-29 1062
2022-12-30 1118
2022-12-31 962
2023-01-01 418
[1460 rows x 1 columns]
--------------
Test statistic: -1.9323134056393474
p-value: 0.31701583612146544
Critical Values: {'1%': -3.434908816804013, '5%': -2.863553406963303, '10%': -2.5678419239852994}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=19477.621, Time=0.32 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=19634.295, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=19613.524, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=19148.783, Time=0.66 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=19632.300, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=19611.593, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=19150.781, Time=1.01 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=19325.037, Time=0.96 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=19610.229, Time=0.16 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=19032.139, Time=2.15 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=18984.822, Time=2.90 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=18882.548, Time=1.71 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=19043.037, Time=1.28 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=19329.113, Time=1.11 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=18878.775, Time=1.92 sec ARIMA(1,1,1)(1,0,1)[6] intercept : AIC=19078.521, Time=1.27 sec ARIMA(1,1,1)(2,0,0)[6] intercept : AIC=19009.155, Time=2.00 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=19009.214, Time=3.27 sec ARIMA(1,1,1)(1,0,0)[6] intercept : AIC=19236.114, Time=0.88 sec ARIMA(1,1,1)(1,0,2)[6] intercept : AIC=18951.249, Time=2.56 sec ARIMA(1,1,0)(2,0,1)[6] intercept : AIC=19294.782, Time=1.65 sec ARIMA(2,1,1)(2,0,1)[6] intercept : AIC=18877.760, Time=2.74 sec ARIMA(2,1,1)(1,0,1)[6] intercept : AIC=18959.123, Time=1.25 sec ARIMA(2,1,1)(2,0,0)[6] intercept : AIC=18890.678, Time=2.58 sec ARIMA(2,1,1)(2,0,2)[6] intercept : AIC=18776.859, Time=3.10 sec ARIMA(2,1,1)(1,0,2)[6] intercept : AIC=18935.821, Time=2.63 sec ARIMA(2,1,0)(2,0,2)[6] intercept : AIC=18619.173, Time=3.06 sec ARIMA(2,1,0)(1,0,2)[6] intercept : AIC=19191.863, Time=1.84 sec ARIMA(2,1,0)(2,0,1)[6] intercept : AIC=18886.949, Time=1.88 sec ARIMA(2,1,0)(1,0,1)[6] intercept : AIC=19328.428, Time=0.45 sec ARIMA(1,1,0)(2,0,2)[6] intercept : AIC=19259.048, Time=1.58 sec ARIMA(2,1,0)(2,0,2)[6] : AIC=18795.376, Time=2.36 sec Best model: ARIMA(2,1,0)(2,0,2)[6] intercept Total fit time: 49.480 seconds Elapsed time: 49.52662450000025
Metrolinx_Stouffville_UNION STATION_Data ARIMA MSE Error: 1850305.838
Metrolinx_Stouffville_UNION STATION_Data ARIMA RMSE Error: 1360.259474
Metrolinx_Stouffville_UNION STATION_Data ARIMA R square: -1.181217675
Model: "sequential_17"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_17 (LSTM) (None, 100) 40800
dense_17 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1065/1065 [==============================] - 6s 5ms/step - loss: 0.0567 Epoch 2/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0481 Epoch 3/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0316 Epoch 4/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0231 Epoch 5/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0161 Epoch 6/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0113 Epoch 7/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0092 Epoch 8/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0096 Epoch 9/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0093 Epoch 10/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0086 Epoch 11/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0084 Epoch 12/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0079 Epoch 13/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0083 Epoch 14/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0079 Epoch 15/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0076 Epoch 16/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0073 Epoch 17/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0070 Epoch 18/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0068 Epoch 19/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0069 Epoch 20/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0068 Epoch 21/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 22/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0071 Epoch 23/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 24/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 25/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 26/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 27/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 28/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 29/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 30/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 31/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 32/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 33/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0057 Epoch 34/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 35/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 36/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 37/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 38/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0067 Epoch 39/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 40/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 41/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 42/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 43/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 44/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 45/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 46/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 47/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 48/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 49/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 50/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 51/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 52/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 53/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 54/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 55/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 56/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 57/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 58/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 59/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 60/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0045 Epoch 61/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0042 Epoch 62/100 1065/1065 [==============================] - 6s 6ms/step - loss: 0.0045 Epoch 63/100 1065/1065 [==============================] - 6s 6ms/step - loss: 0.0041 Epoch 64/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0039 Epoch 65/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0044 Epoch 66/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0039 Epoch 67/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0040 Epoch 68/100 1065/1065 [==============================] - 6s 6ms/step - loss: 0.0055 Epoch 69/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0040 Epoch 70/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0040 Epoch 71/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0037 Epoch 72/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0040 Epoch 73/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0037 Epoch 74/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 75/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 76/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 77/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 78/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 79/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 80/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 81/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 82/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 83/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0041 Epoch 84/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0038 Epoch 85/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 86/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 87/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 88/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0034 Epoch 89/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0035 Epoch 90/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0032 Epoch 91/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 92/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 93/100 1065/1065 [==============================] - 5s 5ms/step - loss: 3588.4934 Epoch 94/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0622 Epoch 95/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0107 Epoch 96/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0072 Epoch 97/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 98/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0058 Epoch 99/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 100/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0050 Elapsed time: 533.3740531999993
Metrolinx_Stouffville_UNION STATION_Data LSTM MSE Error: 1287641.233
Metrolinx_Stouffville_UNION STATION_Data LSTM RMSE Error: 1134.742805
Metrolinx_Stouffville_UNION STATION_Data LSTM R square: -0.5179251769
ds y trips
0 2019-01-03 4499 14
1 2019-01-04 4783 15
2 2019-01-05 4234 15
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Stouffville_UNION STATION_Data #########
18:47:06 - cmdstanpy - INFO - Chain [1] start processing 18:47:07 - cmdstanpy - INFO - Chain [1] done processing
18:47:08 - cmdstanpy - INFO - Chain [1] start processing 18:47:08 - cmdstanpy - INFO - Chain [1] done processing 18:47:09 - cmdstanpy - INFO - Chain [1] start processing 18:47:09 - cmdstanpy - INFO - Chain [1] done processing 18:47:10 - cmdstanpy - INFO - Chain [1] start processing 18:47:11 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.688366099999257
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Stouffville_UNION STATION_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 7756137.304637
rmse 2759.475303
mae 2313.439855
mdape 2.704091
smape 1.80455
coverage 0.342127
dtype: object
ds trips
0 2019-01-03 14
1 2019-01-04 15
2 2019-01-05 15
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1455 2022-12-28 15
1456 2022-12-29 17
1457 2022-12-30 17
1458 2022-12-31 17
1459 2023-01-01 15
[1460 rows x 2 columns]
Metrolinx_Stouffville_UNION STATION_Data Prophet MSE Error: 609233.2062
Metrolinx_Stouffville_UNION STATION_Data Prophet RMSE Error: 780.5339238
Metrolinx_Stouffville_UNION STATION_Data Prophet R square: 0.281810493
=================Metrolinx_Stouffville_OLD ELM GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Stouffville_OLD ELM GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-01-03 4414 9
2019-01-04 4622 9
2019-01-05 4156 9
2019-01-06 0 0
2019-01-07 0 0
Data:
ridership
date
2019-01-03 4414
2019-01-04 4622
2019-01-05 4156
2019-01-06 0
2019-01-07 0
Data Regressors:
trips
date
2019-01-03 9
2019-01-04 9
2019-01-05 9
2019-01-06 0
2019-01-07 0
ridership
date
2019-01-03 4414
2019-01-04 4622
2019-01-05 4156
2019-01-06 0
2019-01-07 0
... ...
2022-12-28 164
2022-12-29 531
2022-12-30 531
2022-12-31 382
2023-01-01 83
[1460 rows x 1 columns]
--------------
Test statistic: -1.890480511309563
p-value: 0.33655254526936074
Critical Values: {'1%': -3.434908816804013, '5%': -2.863553406963303, '10%': -2.5678419239852994}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=19489.819, Time=0.13 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=19640.066, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=19619.291, Time=0.14 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=19136.993, Time=0.66 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=19638.071, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=19617.877, Time=0.06 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=19139.062, Time=1.02 sec ARIMA(0,1,1)(0,0,2)[6] intercept : AIC=19318.585, Time=1.11 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=19616.251, Time=0.16 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=19016.257, Time=2.51 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=19001.028, Time=2.84 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=18874.778, Time=1.73 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=19031.629, Time=1.33 sec ARIMA(0,1,0)(2,0,1)[6] intercept : AIC=19383.582, Time=0.83 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=18941.990, Time=2.09 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=18830.565, Time=1.74 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=18974.940, Time=0.73 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=18929.592, Time=0.86 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=18687.207, Time=3.14 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=18877.540, Time=1.29 sec ARIMA(1,1,2)(2,0,2)[6] intercept : AIC=18897.657, Time=3.40 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=18971.972, Time=2.89 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=18684.603, Time=2.68 sec ARIMA(0,1,2)(1,0,2)[6] : AIC=18877.492, Time=0.62 sec ARIMA(0,1,2)(2,0,1)[6] : AIC=18830.334, Time=1.00 sec ARIMA(0,1,2)(1,0,1)[6] : AIC=18974.415, Time=0.34 sec ARIMA(0,1,1)(2,0,2)[6] : AIC=18680.001, Time=2.23 sec ARIMA(0,1,1)(1,0,2)[6] : AIC=19016.716, Time=1.13 sec ARIMA(0,1,1)(2,0,1)[6] : AIC=18870.480, Time=1.17 sec ARIMA(0,1,1)(1,0,1)[6] : AIC=19138.186, Time=0.56 sec ARIMA(0,1,0)(2,0,2)[6] : AIC=19281.188, Time=0.56 sec ARIMA(1,1,1)(2,0,2)[6] : AIC=18695.334, Time=2.39 sec ARIMA(1,1,0)(2,0,2)[6] : AIC=19248.775, Time=1.35 sec ARIMA(1,1,2)(2,0,2)[6] : AIC=18900.880, Time=2.89 sec Best model: ARIMA(0,1,1)(2,0,2)[6] Total fit time: 45.612 seconds Elapsed time: 45.667472900000575
Metrolinx_Stouffville_OLD ELM GO_Data ARIMA MSE Error: 976278.2566
Metrolinx_Stouffville_OLD ELM GO_Data ARIMA RMSE Error: 988.0679413
Metrolinx_Stouffville_OLD ELM GO_Data ARIMA R square: -0.3773461258
Model: "sequential_18"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_18 (LSTM) (None, 100) 40800
dense_18 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
1065/1065 [==============================] - 6s 5ms/step - loss: 0.0447 Epoch 2/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0275 Epoch 3/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0160 Epoch 4/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0132 Epoch 5/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0106 Epoch 6/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0084 Epoch 7/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0076 Epoch 8/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0069 Epoch 9/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0068 Epoch 10/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0065 Epoch 11/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0068 Epoch 12/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0062 Epoch 13/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0059 Epoch 14/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0060 Epoch 15/100 1065/1065 [==============================] - 6s 6ms/step - loss: 0.0058 Epoch 16/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 17/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 18/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 19/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 20/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 21/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0057 Epoch 22/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 23/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 24/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 25/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 26/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 27/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0050 Epoch 28/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 29/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 30/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 31/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 32/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 33/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 34/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 35/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0040 Epoch 36/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0039 Epoch 37/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 38/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 39/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0039 Epoch 40/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0036 Epoch 41/100 1065/1065 [==============================] - 7s 6ms/step - loss: 0.0039 Epoch 42/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0036 Epoch 43/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 44/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0036 Epoch 45/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0037 Epoch 46/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0034 Epoch 47/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0035 Epoch 48/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0038 Epoch 49/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0034 Epoch 50/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0035 Epoch 51/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0032 Epoch 52/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0037 Epoch 53/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0032 Epoch 54/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 55/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0032 Epoch 56/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0033 Epoch 57/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 58/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0033 Epoch 59/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0031 Epoch 60/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 61/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 62/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0031 Epoch 63/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 64/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0029 Epoch 65/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0033 Epoch 66/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 67/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0031 Epoch 68/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 69/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 70/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0029 Epoch 71/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 72/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0034 Epoch 73/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0031 Epoch 74/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0028 Epoch 75/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 76/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0031 Epoch 77/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0028 Epoch 78/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0030 Epoch 79/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0026 Epoch 80/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0025 Epoch 81/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0030 Epoch 82/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0028 Epoch 83/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0023 Epoch 84/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 85/100 1065/1065 [==============================] - 6s 6ms/step - loss: 0.0025 Epoch 86/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0025 Epoch 87/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0027 Epoch 88/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0026 Epoch 89/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0023 Epoch 90/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0026 Epoch 91/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0024 Epoch 92/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0026 Epoch 93/100 1065/1065 [==============================] - 5s 5ms/step - loss: 0.0024 Epoch 94/100 1065/1065 [==============================] - 6s 6ms/step - loss: 0.0028 Epoch 95/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0028 Epoch 96/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0025 Epoch 97/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0027 Epoch 98/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0024 Epoch 99/100 1065/1065 [==============================] - 6s 6ms/step - loss: 0.0026 Epoch 100/100 1065/1065 [==============================] - 6s 5ms/step - loss: 0.0025 Elapsed time: 551.0871349999998
Metrolinx_Stouffville_OLD ELM GO_Data LSTM MSE Error: 1395320.625
Metrolinx_Stouffville_OLD ELM GO_Data LSTM RMSE Error: 1181.236905
Metrolinx_Stouffville_OLD ELM GO_Data LSTM R square: -0.9685365764
ds y trips
0 2019-01-03 4414 9
1 2019-01-04 4622 9
2 2019-01-05 4156 9
3 2019-01-06 0 0
4 2019-01-07 0 0
########## Cross validation for Metrolinx_Stouffville_OLD ELM GO_Data #########
18:57:25 - cmdstanpy - INFO - Chain [1] start processing 18:57:25 - cmdstanpy - INFO - Chain [1] done processing
18:57:26 - cmdstanpy - INFO - Chain [1] start processing 18:57:26 - cmdstanpy - INFO - Chain [1] done processing 18:57:27 - cmdstanpy - INFO - Chain [1] start processing 18:57:28 - cmdstanpy - INFO - Chain [1] done processing 18:57:29 - cmdstanpy - INFO - Chain [1] start processing 18:57:29 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 5.562276900000143
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Stouffville_OLD ELM GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 2734833.909382
rmse 1625.729715
mae 1327.079163
mdape 2.043445
smape 1.487796
coverage 0.451112
dtype: object
ds trips
0 2019-01-03 9
1 2019-01-04 9
2 2019-01-05 9
3 2019-01-06 0
4 2019-01-07 0
... ... ...
1455 2022-12-28 3
1456 2022-12-29 6
1457 2022-12-30 6
1458 2022-12-31 6
1459 2023-01-01 3
[1460 rows x 2 columns]
Metrolinx_Stouffville_OLD ELM GO_Data Prophet MSE Error: 484734.4359
Metrolinx_Stouffville_OLD ELM GO_Data Prophet RMSE Error: 696.2287238
Metrolinx_Stouffville_OLD ELM GO_Data Prophet R square: 0.3161303217
=================Metrolinx_Stouffville_MOUNT JOY GO_Data==================
C:\Users\DKici\Desktop\Metrolinx\notebook\ts_data\Metrolinx_Stouffville_MOUNT JOY GO_Data.csv
time series loaded!
Data All:
ridership trips
date
2019-04-09 196 5
2019-04-10 169 3
2019-04-11 229 5
2019-04-12 213 5
2019-04-13 213 5
Data:
ridership
date
2019-04-09 196
2019-04-10 169
2019-04-11 229
2019-04-12 213
2019-04-13 213
Data Regressors:
trips
date
2019-04-09 5
2019-04-10 3
2019-04-11 5
2019-04-12 5
2019-04-13 5
ridership
date
2019-04-09 196
2019-04-10 169
2019-04-11 229
2019-04-12 213
2019-04-13 213
... ...
2022-12-28 875
2022-12-29 657
2022-12-30 745
2022-12-31 692
2023-01-01 685
[1364 rows x 1 columns]
--------------
Test statistic: -1.2417555735943089
p-value: 0.6553806861518309
Critical Values: {'1%': -3.4352394560472033, '5%': -2.863699314925706, '10%': -2.567919624081087}
Weak evidence against the null hypothesis
Fail to reject the null hypothesis
Data has a unit root and is non-stationary
C:\Users\DKici\Desktop\Metrolinx\notebook\utils.py:248: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure. data_decompose_add.plot().show()
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
C:\Users\DKici\AppData\Roaming\Python\Python37\site-packages\matplotlib\axes\_base.py:2475: UserWarning: Warning: converting a masked element to nan.
xys = np.asarray(xys)
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\statsmodels\graphics\tsaplots.py:353: FutureWarning: The default method 'yw' can produce PACF values outside of the [-1,1] interval. After 0.13, the default will change tounadjusted Yule-Walker ('ywm'). You can use this method now by setting method='ywm'.
FutureWarning,
Performing stepwise search to minimize aic ARIMA(0,1,0)(1,0,1)[6] intercept : AIC=11836.174, Time=0.30 sec ARIMA(0,1,0)(0,0,0)[6] intercept : AIC=11863.575, Time=0.02 sec ARIMA(1,1,0)(1,0,0)[6] intercept : AIC=11857.046, Time=0.12 sec ARIMA(0,1,1)(0,0,1)[6] intercept : AIC=11718.013, Time=0.37 sec ARIMA(0,1,0)(0,0,0)[6] : AIC=11861.575, Time=0.01 sec ARIMA(0,1,1)(0,0,0)[6] intercept : AIC=11762.793, Time=0.15 sec ARIMA(0,1,1)(1,0,1)[6] intercept : AIC=11717.302, Time=0.57 sec ARIMA(0,1,1)(1,0,0)[6] intercept : AIC=11730.469, Time=0.32 sec ARIMA(0,1,1)(2,0,1)[6] intercept : AIC=11622.848, Time=0.84 sec ARIMA(0,1,1)(2,0,0)[6] intercept : AIC=11693.320, Time=0.81 sec ARIMA(0,1,1)(2,0,2)[6] intercept : AIC=11536.399, Time=2.81 sec ARIMA(0,1,1)(1,0,2)[6] intercept : AIC=11653.857, Time=1.27 sec ARIMA(0,1,0)(2,0,2)[6] intercept : AIC=11776.941, Time=1.82 sec ARIMA(1,1,1)(2,0,2)[6] intercept : AIC=inf, Time=2.82 sec ARIMA(0,1,2)(2,0,2)[6] intercept : AIC=11513.368, Time=2.10 sec ARIMA(0,1,2)(1,0,2)[6] intercept : AIC=11510.087, Time=0.86 sec ARIMA(0,1,2)(0,0,2)[6] intercept : AIC=11525.318, Time=0.98 sec ARIMA(0,1,2)(1,0,1)[6] intercept : AIC=11520.346, Time=0.75 sec ARIMA(0,1,2)(0,0,1)[6] intercept : AIC=11533.792, Time=0.48 sec ARIMA(0,1,2)(2,0,1)[6] intercept : AIC=11503.919, Time=1.15 sec ARIMA(0,1,2)(2,0,0)[6] intercept : AIC=11525.466, Time=0.85 sec ARIMA(0,1,2)(1,0,0)[6] intercept : AIC=11535.729, Time=0.39 sec ARIMA(1,1,2)(2,0,1)[6] intercept : AIC=11505.890, Time=1.48 sec ARIMA(1,1,1)(2,0,1)[6] intercept : AIC=11535.235, Time=1.27 sec ARIMA(0,1,2)(2,0,1)[6] : AIC=11501.987, Time=0.33 sec ARIMA(0,1,2)(1,0,1)[6] : AIC=11518.362, Time=0.42 sec ARIMA(0,1,2)(2,0,0)[6] : AIC=11523.487, Time=0.31 sec ARIMA(0,1,2)(2,0,2)[6] : AIC=11511.412, Time=1.46 sec ARIMA(0,1,2)(1,0,0)[6] : AIC=11533.746, Time=0.16 sec ARIMA(0,1,2)(1,0,2)[6] : AIC=11508.138, Time=0.39 sec ARIMA(0,1,1)(2,0,1)[6] : AIC=11620.909, Time=0.40 sec ARIMA(1,1,2)(2,0,1)[6] : AIC=11503.942, Time=0.64 sec ARIMA(1,1,1)(2,0,1)[6] : AIC=11533.299, Time=0.60 sec Best model: ARIMA(0,1,2)(2,0,1)[6] Total fit time: 27.274 seconds Elapsed time: 27.33434390000002
Metrolinx_Stouffville_MOUNT JOY GO_Data ARIMA MSE Error: 122998.0702
Metrolinx_Stouffville_MOUNT JOY GO_Data ARIMA RMSE Error: 350.7108071
Metrolinx_Stouffville_MOUNT JOY GO_Data ARIMA R square: -0.5472177017
Model: "sequential_19"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm_19 (LSTM) (None, 100) 40800
dense_19 (Dense) (None, 1) 101
=================================================================
Total params: 40,901
Trainable params: 40,901
Non-trainable params: 0
_________________________________________________________________
Epoch 1/100
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\ipykernel_launcher.py:101: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
969/969 [==============================] - 6s 5ms/step - loss: 0.0161 Epoch 2/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0144 Epoch 3/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0139 Epoch 4/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0133 Epoch 5/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0132 Epoch 6/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0120 Epoch 7/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0107 Epoch 8/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0104 Epoch 9/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0101 Epoch 10/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0096 Epoch 11/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0092 Epoch 12/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0086 Epoch 13/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0083 Epoch 14/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0082 Epoch 15/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0079 Epoch 16/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0074 Epoch 17/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0073 Epoch 18/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0075 Epoch 19/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0071 Epoch 20/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0071 Epoch 21/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0070 Epoch 22/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0068 Epoch 23/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 24/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 25/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0066 Epoch 26/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 27/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 28/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0064 Epoch 29/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 30/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0062 Epoch 31/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0063 Epoch 32/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 33/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0062 Epoch 34/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0063 Epoch 35/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0061 Epoch 36/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0058 Epoch 37/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 38/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 39/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 40/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 41/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 42/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0059 Epoch 43/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0057 Epoch 44/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0057 Epoch 45/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0060 Epoch 46/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0057 Epoch 47/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0057 Epoch 48/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 49/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 50/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0059 Epoch 51/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 52/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0057 Epoch 53/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0058 Epoch 54/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0056 Epoch 55/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0056 Epoch 56/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0074 Epoch 57/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0061 Epoch 58/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0057 Epoch 59/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0054 Epoch 60/100 969/969 [==============================] - 7s 7ms/step - loss: 0.0055 Epoch 61/100 969/969 [==============================] - 8s 8ms/step - loss: 0.0053 Epoch 62/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0059 Epoch 63/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 64/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0055 Epoch 65/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 66/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0054 Epoch 67/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 68/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0053 Epoch 69/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0052 Epoch 70/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0051 Epoch 71/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0051 Epoch 72/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 73/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0052 Epoch 74/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 75/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0050 Epoch 76/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 77/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 78/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 79/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 80/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0049 Epoch 81/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0049 Epoch 82/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0049 Epoch 83/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 84/100 969/969 [==============================] - 6s 6ms/step - loss: 0.0048 Epoch 85/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0050 Epoch 86/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0047 Epoch 87/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0046 Epoch 88/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 89/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0048 Epoch 90/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0045 Epoch 91/100 969/969 [==============================] - 5s 6ms/step - loss: 0.0047 Epoch 92/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0049 Epoch 93/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0043 Epoch 94/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 95/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0044 Epoch 96/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 97/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0046 Epoch 98/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0045 Epoch 99/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0042 Epoch 100/100 969/969 [==============================] - 5s 5ms/step - loss: 0.0046 Elapsed time: 528.9143233000013
Metrolinx_Stouffville_MOUNT JOY GO_Data LSTM MSE Error: 253094.0213
Metrolinx_Stouffville_MOUNT JOY GO_Data LSTM RMSE Error: 503.0845072
Metrolinx_Stouffville_MOUNT JOY GO_Data LSTM R square: -2.18372109
ds y trips
0 2019-04-09 196 5
1 2019-04-10 169 3
2 2019-04-11 229 5
3 2019-04-12 213 5
4 2019-04-13 213 5
########## Cross validation for Metrolinx_Stouffville_MOUNT JOY GO_Data #########
19:07:02 - cmdstanpy - INFO - Chain [1] start processing 19:07:02 - cmdstanpy - INFO - Chain [1] done processing
19:07:03 - cmdstanpy - INFO - Chain [1] start processing 19:07:03 - cmdstanpy - INFO - Chain [1] done processing 19:07:05 - cmdstanpy - INFO - Chain [1] start processing 19:07:05 - cmdstanpy - INFO - Chain [1] done processing
Elapsed time: 4.3020829999986745
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
C:\Users\DKici\Anaconda3\envs\Dart_env_37\lib\site-packages\prophet\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.
x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])
Ridership Metrolinx_Stouffville_MOUNT JOY GO_Data- CV-metrics mean
horizon 201 days 00:00:00
mse 50450.117086
rmse 209.785902
mae 158.528756
mdape 0.446724
smape 0.721066
coverage 0.326581
dtype: object
ds trips
0 2019-04-09 5
1 2019-04-10 3
2 2019-04-11 5
3 2019-04-12 5
4 2019-04-13 5
... ... ...
1359 2022-12-28 12
1360 2022-12-29 10
1361 2022-12-30 10
1362 2022-12-31 10
1363 2023-01-01 12
[1364 rows x 2 columns]
Metrolinx_Stouffville_MOUNT JOY GO_Data Prophet MSE Error: 23166.66539
Metrolinx_Stouffville_MOUNT JOY GO_Data Prophet RMSE Error: 152.2059966
Metrolinx_Stouffville_MOUNT JOY GO_Data Prophet R square: 0.7085818116
my_evaluation.sort_values(by = 'R2', ascending=False)
| model | ts | mse | rmse | R2 | time (min) | |
|---|---|---|---|---|---|---|
| 32 | Prophet | Metrolinx_Kitchener_MOUNT PLEASANT GO_Data | 2.830351e+04 | 168.236482 | 0.783950 | 0.094075 |
| 29 | Prophet | Metrolinx_Kitchener_KITCHENER GO_Data | 4.286345e+05 | 654.701833 | 0.772813 | 0.106755 |
| 5 | Prophet | Metrolinx_Barrie_AURORA GO_Data | 1.165693e+04 | 107.967280 | 0.741255 | 0.095547 |
| 20 | Prophet | Metrolinx_Lakeshore West_ALDERSHOT GO_Data | 1.391345e+06 | 1179.552805 | 0.728033 | 0.094173 |
| 50 | Prophet | Metrolinx_Stouffville_MOUNT JOY GO_Data | 2.316667e+04 | 152.205997 | 0.708582 | 0.071701 |
| 26 | Prophet | Metrolinx_Kitchener_UNION STATION_Data | 8.612845e+05 | 928.054167 | 0.701372 | 0.090264 |
| 14 | Prophet | Metrolinx_Lakeshore East_OSHAWA GO_Data | 3.716981e+06 | 1927.947374 | 0.557740 | 0.075365 |
| 2 | Prophet | Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data | 5.500281e+05 | 741.638781 | 0.521176 | 0.083982 |
| 17 | Prophet | Metrolinx_Lakeshore West_UNION STATION_Data | 8.146776e+06 | 2854.255848 | 0.501102 | 0.085385 |
| 11 | Prophet | Metrolinx_Lakeshore East_UNION STATION_Data | 3.449942e+06 | 1857.401976 | 0.493546 | 0.095676 |
| 23 | Prophet | Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data | 3.095833e+05 | 556.402067 | 0.462330 | 0.086622 |
| 35 | Prophet | Metrolinx_Milton_UNION STATION_Data | 8.788269e+05 | 937.457670 | 0.417350 | 0.088303 |
| 38 | Prophet | Metrolinx_Milton_MILTON GO_Data | 1.040369e+06 | 1019.984776 | 0.405924 | 0.093610 |
| 8 | Prophet | Metrolinx_Barrie_UNION STATION_Data | 8.094233e+05 | 899.679537 | 0.361494 | 0.091949 |
| 47 | Prophet | Metrolinx_Stouffville_OLD ELM GO_Data | 4.847344e+05 | 696.228724 | 0.316130 | 0.092705 |
| 44 | Prophet | Metrolinx_Stouffville_UNION STATION_Data | 6.092332e+05 | 780.533924 | 0.281810 | 0.094806 |
| 41 | Prophet | Metrolinx_Richmond Hill_UNION STATION_Data | 1.855332e+05 | 430.735661 | -0.131001 | 0.088551 |
| 33 | ARIMA | Metrolinx_Milton_UNION STATION_Data | 1.818460e+06 | 1348.503054 | -0.205613 | 0.673857 |
| 24 | ARIMA | Metrolinx_Kitchener_UNION STATION_Data | 3.597640e+06 | 1896.744515 | -0.247389 | 0.780045 |
| 36 | ARIMA | Metrolinx_Milton_MILTON GO_Data | 2.189601e+06 | 1479.730134 | -0.250315 | 0.648336 |
| 12 | ARIMA | Metrolinx_Lakeshore East_OSHAWA GO_Data | 1.060020e+07 | 3255.794832 | -0.261250 | 0.582783 |
| 39 | ARIMA | Metrolinx_Richmond Hill_UNION STATION_Data | 2.070535e+05 | 455.031312 | -0.262188 | 1.007997 |
| 45 | ARIMA | Metrolinx_Stouffville_OLD ELM GO_Data | 9.762783e+05 | 988.067941 | -0.377346 | 0.761125 |
| 43 | LSTM | Metrolinx_Stouffville_UNION STATION_Data | 1.287641e+06 | 1134.742805 | -0.517925 | 8.889568 |
| 48 | ARIMA | Metrolinx_Stouffville_MOUNT JOY GO_Data | 1.229981e+05 | 350.710807 | -0.547218 | 0.455572 |
| 37 | LSTM | Metrolinx_Milton_MILTON GO_Data | 2.983731e+06 | 1727.348092 | -0.703782 | 8.755329 |
| 28 | LSTM | Metrolinx_Kitchener_KITCHENER GO_Data | 3.341113e+06 | 1827.871233 | -0.770877 | 8.338663 |
| 0 | ARIMA | Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data | 2.055633e+06 | 1433.747861 | -0.789520 | 1.200123 |
| 3 | ARIMA | Metrolinx_Barrie_AURORA GO_Data | 8.427999e+04 | 290.310165 | -0.870735 | 0.993617 |
| 22 | LSTM | Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data | 1.088855e+06 | 1043.482214 | -0.891072 | 7.214860 |
| 31 | LSTM | Metrolinx_Kitchener_MOUNT PLEASANT GO_Data | 2.501167e+05 | 500.116724 | -0.909227 | 8.092103 |
| 40 | LSTM | Metrolinx_Richmond Hill_UNION STATION_Data | 3.219042e+05 | 567.365991 | -0.962311 | 8.501961 |
| 46 | LSTM | Metrolinx_Stouffville_OLD ELM GO_Data | 1.395321e+06 | 1181.236905 | -0.968537 | 9.184786 |
| 42 | ARIMA | Metrolinx_Stouffville_UNION STATION_Data | 1.850306e+06 | 1360.259474 | -1.181218 | 0.825444 |
| 34 | LSTM | Metrolinx_Milton_UNION STATION_Data | 3.419331e+06 | 1849.143299 | -1.266968 | 8.173794 |
| 25 | LSTM | Metrolinx_Kitchener_UNION STATION_Data | 6.772796e+06 | 2602.459632 | -1.348293 | 7.446252 |
| 21 | ARIMA | Metrolinx_Lakeshore West_HAMILTON GO CENTRE_Data | 1.378594e+06 | 1174.135607 | -1.394277 | 0.843471 |
| 6 | ARIMA | Metrolinx_Barrie_UNION STATION_Data | 3.088077e+06 | 1757.292587 | -1.436000 | 0.824528 |
| 18 | ARIMA | Metrolinx_Lakeshore West_ALDERSHOT GO_Data | 1.265126e+07 | 3556.861530 | -1.472952 | 0.579358 |
| 15 | ARIMA | Metrolinx_Lakeshore West_UNION STATION_Data | 4.321167e+07 | 6573.558168 | -1.646225 | 0.516364 |
| 19 | LSTM | Metrolinx_Lakeshore West_ALDERSHOT GO_Data | 1.380715e+07 | 3715.796713 | -1.698892 | 7.507331 |
| 16 | LSTM | Metrolinx_Lakeshore West_UNION STATION_Data | 4.852517e+07 | 6966.000678 | -1.971617 | 7.182725 |
| 1 | LSTM | Metrolinx_Barrie_ALLANDALE WATERFRONT GO_Data | 3.450533e+06 | 1857.561151 | -2.003843 | 6.940892 |
| 9 | ARIMA | Metrolinx_Lakeshore East_UNION STATION_Data | 2.086968e+07 | 4568.334825 | -2.063683 | 1.136466 |
| 27 | ARIMA | Metrolinx_Kitchener_KITCHENER GO_Data | 5.939523e+06 | 2437.113618 | -2.148101 | 0.493863 |
| 30 | ARIMA | Metrolinx_Kitchener_MOUNT PLEASANT GO_Data | 4.168348e+05 | 645.627444 | -2.181843 | 1.187052 |
| 49 | LSTM | Metrolinx_Stouffville_MOUNT JOY GO_Data | 2.530940e+05 | 503.084507 | -2.183721 | 8.815239 |
| 4 | LSTM | Metrolinx_Barrie_AURORA GO_Data | 1.595617e+05 | 399.451775 | -2.541738 | 6.982528 |
| 7 | LSTM | Metrolinx_Barrie_UNION STATION_Data | 5.683113e+06 | 2383.927967 | -3.483069 | 7.161308 |
| 10 | LSTM | Metrolinx_Lakeshore East_UNION STATION_Data | 3.510692e+07 | 5925.109344 | -4.153718 | 7.046959 |
| 13 | LSTM | Metrolinx_Lakeshore East_OSHAWA GO_Data | 4.649342e+07 | 6818.608016 | -4.531954 | 6.613414 |
my_evaluation.to_csv(f"{output_path}\\results\\Model_results_Stations.csv")
As a result, we received acceptable results for following 6 stations with Prophet model:
Metrolinx_Kitchener_MOUNT PLEASANT GO_Data
Metrolinx_Kitchener_KITCHENER GO_Data
Metrolinx_Barrie_AURORA GO_Data
Metrolinx_Lakeshore West_ALDERSHOT GO_Data
Metrolinx_Stouffville_MOUNT JOY GO_Data
Metrolinx_Kitchener_UNION STATION_Data